Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2007-06-04

00:00:28 Quit Rondom (Read error: 113 (No route to host))
00:01:26 Join Rondom [0] (
00:03:19 Part TrueJournals
00:03:53 Quit petur ("zzz")
00:10:47 Join ducbian_ [0] (
00:18:49 Join Rob222241 [0] (
00:22:08 Join perldiver [0] (
00:23:34 Join TrueJournals [0] (
00:24:34 Part n1s
00:28:30 Quit ducbian (Read error: 110 (Connection timed out))
00:30:27 Quit pearldiver (Read error: 110 (Connection timed out))
00:30:33 Quit Rondom ("Ex-Chat")
00:32:09 Quit Buschel ()
00:36:24 Quit Rob2222 (Read error: 110 (Connection timed out))
00:38:22DerPapstgood night at all
00:38:26 Part TrueJournals
00:38:57 Quit DerPapst ("So Long And Thanks For All The Fish!")
00:39:40 Join lxx [0] (i=lexx@
00:40:30lxxwhat the best prog for upload mp3`s to my ipod?
00:41:35 Quit Arathis ("Bye, bye")
00:43:25element_glxx: depends, which OS are you running? and what are you looking to do? sync, or just transfer?
00:45:04 Join Davide-NYC [0] (
00:45:38 Quit Davide-NYC (Client Quit)
00:47:14lxxi use windows
00:47:24***Saving seen data "./dancer.seen"
00:47:25lxxfor transfer
00:49:03linuxstbYou want something different to simply copying them in windows explorer?
00:50:06 Quit ompaul ("bye")
00:50:55saratogalxx: you could use foobar's copy +rename feature
00:51:29lxxbut i ise now official firware
00:51:43lxxi need prof for official firmware
00:52:15element_glxx: you want to run a dual boot iPod?
00:53:57 Quit ender` (" I think I remember an episode of MacGyver where he overthrew a violent dictator with a rubber band, 2 bottle caps, and some )
00:54:09linuxstblxx: This isn't a support forum for Apple's firmware...
00:54:21element_glxx: if you need dual boot, then any program that writes to the itunes DB will do (itunes, winamp, mediamonkey)
00:55:18lxx <linuxstb> where i can ask about id?
00:56:18 Part krush1704
00:56:46lxxrockbox support fm transmitters?
00:57:32linuxstbSee this page for Rockbox's compatibilty with ipod accessories:
00:58:23 Join krush1704 [0] (
00:58:36 Join Rick| [0] (n=rhelmus@
00:59:02 Quit spiorf ("Read error: 110 (Connection timed out)")
00:59:41lxxothers fm tranmiters dont tested or just not working?
01:01:02linuxstbNot tested. Or at least, not added to the wiki page.
01:01:21 Quit bluebrother ("leaving")
01:01:30 Join jhulst [0] (
01:02:37lxxbut what the best fm transmitter?
01:02:55Rick|I wonder if anyone could help me with my ipod nano not wanting to (re)boot anymore?
01:05:39Nico_Pfrom : "Support for RockBox-based devices in Amarok." anyone know about this ?
01:05:46linuxstbRick|: What's the problem?
01:05:51 Quit Entasis (Read error: 104 (Connection reset by peer))
01:06:13Rick|well it seems that it happened after the battery was fully discharged
01:07:47Rick|I've recharged for a few hours, though everything seems to be dead; blank screen, ipod not recognized by pc, not possible to reboot (tried to toggling hold key)
01:08:01saratogapretty much all fm transmitters are crap
01:08:13saratogaFM is not a very good medium for sending audio
01:08:23saratogaparticularly at low power
01:08:48element_gsaratoga: agreed, aux input car decks are much better
01:08:59linuxstbRick|: How long have you been holding MENU+SELECT together? Some people need 30+ seconds.
01:09:31 Quit Siltaar (Remote closed the connection)
01:09:52lxxbut i whant find good transmitter with ipod charger and i dont have others options to connect ipod in car
01:10:04Rick|several times for about 10-20 and a few 30-40
01:10:12 Quit DC1 (Read error: 110 (Connection timed out))
01:10:28Rick|guess it won't hurt to try it another time for about a minute
01:13:12Rick|nope 2 times ~1 min and still very sleepy
01:14:43Rick|I wonder if Rockbox could anything have to do with it? It went offline while booting RB
01:22:17 Part element_g ("Kopete 0.12.4 :")
01:23:11 Part krush1704
01:28:02lxxso anyone here use fm transmitter in car?
01:29:15scorcheRick|: not really...i would suggest that you leave it plugged in overnight, and then try it again
01:32:48Rick|scorche: ok...I'll try that and see if it's tomorrow any better
01:33:27midkayRick|: i've had this a few times before. what i've actually found is that it only seems to fix itself if you leave it *unplugged* for like a full day.
01:33:48midkayperhaps as in, drain the battery...
01:34:22midkayhad the problem the first time, charged it all day on and off, no fix.. left it sitting for a day, it started working again. happened again a few days ago. left it unplugged for a day, started working again.
01:34:37Rick|than it might be longer, since it has been plugged in for a few hours already (assuming it actually charged)
01:34:59midkayRick|: well, i surely charged it "fully" all day before leaving it unplugged for a day.
01:35:03 Quit obo ("bye")
01:35:09Rick|ah cool :)
01:35:13midkaystill seemed to fix itself.. so *shrugs*
01:35:16midkaythat's my solution :)
01:35:37Rick|well I give it some rest then :)
01:35:47midkayRick|: that should help. :)
01:36:26 Quit Nico_P (Remote closed the connection)
01:36:36Rick|how did you found out it worked again? just plugged it in after a day?
01:36:41 Join BHSPitMonkey [0] (
01:37:49midkayRick: i left it sitting with the hold switch on for probably 18 hours, only tried it a few times.. and then i turned the switch off once at the end and pressed MENU, the screen flickered on for a few seconds before dying.
01:37:53midkaythe battery was completely dead.
01:38:20midkayseems like you need to somehow let it discharge during that time before it'll come back on...
01:38:38midkayi don't understand what causes it or why menu+select won't reset it, but it's happened twice and that's worked both times. quite annoying.
01:39:40 Quit ducbian_ (Read error: 110 (Connection timed out))
01:40:23Rick|did it happen when your battery was nearly drained aswell?
01:40:38midkayi don't remember. but i definitely charged it fully before leaving it unplugged that period of time.
01:40:38 Quit Daishi ("Client exiting...")
01:40:48midkayso it's like it's stuck on, draining the battery as if it were running...
01:40:53midkayit wouldn't have died that quickly if it were "off".
01:41:01 Quit XavierGr ("One firmware to rule them all!")
01:41:20Rick|oh well I hope it works :)
01:41:37 Part pixelma
01:41:40midkayyeah, good luck. gotta run :)
01:41:56Rick|ok thanks so far :)
01:45:05 Quit lxx ()
01:45:38 Join webguest87 [0] (i=4a841a2a@gateway/web/cgi-irc/
01:46:19webguest87 how can you guys see peoples ip adresses?
01:47:55scorchehelo again...
01:49:49scorcheplease dont spam the channel....the guidelines can be located in the topic
01:51:25webguest87Yes sir. Can i get my history removed? My brother got on whilst i was getting lunch. He is 13. And immature. Sorry...
01:51:49scorchethe logs are the logs, and they dont get edited
01:52:26scorchebut, for the purposes of this channel, i will not consider you as a spammer, unless it happens again
01:54:47webguest87I apoligize for the inconviences of my brothers doings. I swear not to do it again.
01:55:05scorchewell, if it was your brother, then you didnt do it before, no? ;)
01:55:43 Part Rick|
01:57:05webguest87You mispelled hello during your first post...
01:57:17scorchei know...i make a lot of typos
01:57:28scorcheso many, that i have stopped correcting them largely
01:58:13webguest87If no one is discussing anything, what do you guys do?
01:58:38scorchewell, this isnt a social channel, so it sits silent until needed
02:01:15webguest87OK. cool. I have a question. Where can I get the WAD files for Doom to work? I downloaded one and I cannot play any of my cool plugins?
02:01:36webguest87I hope they are free...
02:06:27webguest87Is anyone else on besides me and scorche?
02:08:58saratogastupid question, how do i format a number with printf so that it gets zero padded if its small?
02:09:24saratogaI want to print with fixed widths for fp numbers
02:09:31saratoga"%+5.4lf" isn't doing it
02:10:18webguest87stupid question, why is there a setting that makes the screen display upside-down?
02:12:10saratogaoh never mind
02:12:58 Quit lids (Remote closed the connection)
02:15:59 Quit webguest87 ("CGI:IRC")
02:16:19 Join webguest87 [0] (i=4a841a2a@gateway/web/cgi-irc/
02:20:26 Join lids [0] (i=lds@gateway/tor/x-20e514624ee1eb89)
02:22:07 Join TrueJournals [0] (
02:24:45 Quit webguest87 ("CGI:IRC (EOF)")
02:28:20 Quit entheh ("^~")
02:36:16 Join XiLe [0] (i=47ea5344@gateway/web/cgi-irc/
02:38:20 Quit XiLe (Client Quit)
02:40:10 Part TrueJournals
02:46:32 Join toffe82 [0] (n=chatzill@
02:47:20 Join krazykit [0] (
02:47:26***Saving seen data "./dancer.seen"
02:47:49 Join TrueJournals [0] (
02:50:03 Join XiLe [0] (i=47ea5344@gateway/web/cgi-irc/
02:51:37XiLeare there any ways to buy music other than cd that i can play on rockbox? allofmp3 isnt accepting credit cards i guess
02:55:10saratogaiTunes, provided you buy hte tracks without DRM
02:55:36XiLeanything for the major record labels?
02:56:41XiLei heard emusic is only independent stuff, and itunes is only emi records :( am i best off just buying the cd, quality aside?
02:57:12 Quit BigMac (Remote closed the connection)
02:58:02saratogaitunes is supposed to be getting most of the other records to go DRM-less eventually
02:58:03 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
02:58:12saratogaso i guess you could buy CDs for now
02:58:25XiLemmm okay thanks
02:59:17chrisjs169you always got google...or are you looking for a more legal way?
02:59:32XiLeim looking legal
02:59:50XiLeive been using google with intitle and such for too long
03:11:25 Quit XiLe ("CGI:IRC (EOF)")
03:17:09 Quit jhulst (Read error: 113 (No route to host))
03:41:27 Join Trih`Xeem [0] (
03:54:42chrisjs169what does Can't load from partition bad checksum mean?
03:55:06chrisjs169Well, I know it means that the checksum is bad, but what would be required to fix it?
03:59:31 Part TrueJournals
04:05:56 Quit sbeh (
04:06:22NJoinsbeh [0] (
04:07:03 Quit Trih`Xeem (Read error: 110 (Connection timed out))
04:32:54 Quit YouCeyE (Read error: 104 (Connection reset by peer))
04:34:51 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
04:45:50alienbiker99has there been an ignore "the" patch added to the svn?
04:47:27***Saving seen data "./dancer.seen"
05:02:46 Join Brunellus [0] (n=luigi@unaffiliated/brunellus)
05:02:56 Part Brunellus ("Ex-Chat")
05:04:46saratogaalienbiker99: no, theres a lot of disagreement about how to implement that feature
05:05:09saratogai think people dislike committing english specific patches, since rockbox supports many languages
05:06:41 Join secleinteer [0] (n=scl@
05:06:56alienbiker99ah ok, actually i think i saw one a while back supporting like german and other but oh well
05:17:45 Quit saratoga ("CGI:IRC (EOF)")
05:19:10 Join tradotto_ [0] (
05:20:29 Quit kkurbjun (Read error: 110 (Connection timed out))
05:21:53 Quit tradotto_ (Remote closed the connection)
05:38:38Nimdaeapparently i've only been idle for 2 days
05:38:50Nimdaei know i've been idle longer than that...what gives
05:39:11Nimdaeso, i have an interesting bug, about to searchf or it on the tracker
05:51:06 Join tradotto [0] (
06:11:59 Quit rotator ()
06:13:09 Join printfXh4 [0] (
06:32:52 Join nick89 [0] (
06:33:13 Quit nick89 (Client Quit)
06:39:00 Part toffe82
06:47:31***Saving seen data "./dancer.seen"
06:52:13 Join HellDragon [0] (i=JD@unaffiliated/helldragon)
07:01:02 Join NIXON-Trent [0] (
07:23:46 Join aliask [0] (
07:32:47 Join ducbian [0] (
07:37:16 Join Brunellus [0] (n=luigi@unaffiliated/brunellus)
07:43:12 Quit tradotto (Read error: 104 (Connection reset by peer))
07:45:12 Join _Veseliq_ [0] (
07:48:16 Join Rob2222 [0] (
07:50:42 Part maffe
08:01:16 Join Siltaar [0] (n=Siltaar@
08:06:36 Quit Rob222241 (Read error: 110 (Connection timed out))
08:08:33 Join pondlife [0] (
08:10:17 Quit lids (Remote closed the connection)
08:11:07NIXON-Trentanyone here ?
08:16:25pondlifeI am
08:19:35 Join Rob222241 [0] (
08:20:49NIXON-Trentah u wouldnt know much about linuxipods movie playback ?
08:22:05SiltaarNIXON-Trent, I know that it has been discussed a lot recently
08:22:39Siltaarso it should'nt have already been implemented
08:22:58Siltaarspecially if the related user manual don't mention it
08:23:03Siltaardo it ?
08:23:45 Join ender` [0] (
08:26:03NIXON-Trenti've been trying to find out if theirs can seek (fast forward and rewind)
08:26:20NIXON-Trentas its really annoying with the rockbox player as if u dont sit and watch the whole movie.. there is no point :/
08:26:49NIXON-Trentas im often traveling on trains and so forth its annoying as i have to turn it off then back on and loose my place in the movie
08:28:34 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:32:03 Quit Rob2222 (Read error: 110 (Connection timed out))
08:35:37SiltaarNIXON-Trent, I understand
08:36:00Siltaarat least a cold resum function would be enough
08:37:06NIXON-Trentso i was just seeing if the ipodlinux player supported some control options ...
08:37:31NIXON-Trenteven though i do like rockbox better i'd just run linux on the side just for the player feature..
08:38:42 Join lids [0] (i=lds@gateway/tor/x-9d722bde199b5976)
08:46:50 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
08:47:34***Saving seen data "./dancer.seen"
08:49:21 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
08:49:53linuxstb_NIXON-Trent: Which ipod do you have?
08:55:37pondlifeamiconn: When you get time, please could you look at and see if this changes the .talk clip configuration sensibly?
08:58:03amiconnlinuxstb: Btw, even an .rvf player will have the stuttering problem on ipods as long as the grayscale lib runs on core0 and no improvements regarding the lcd controller i/f can be made
08:58:36*GodEater is watching Linus Torvald's speed on source code management here : : Quite interesting.
08:58:52amiconnOr rather, .rvf has its own "grayscale lib" mechanism, but the effect is the same
08:59:09amiconnpondlife: Will do...
08:59:23amiconn(may need a reminder)
09:01:00pondlifeamiconn: No problem.. It just seperates out the checking of .talk clips, allowing a fallback to the user's choice of spelling/numbers/none
09:01:00linuxstb_amiconn: Did you try that quick change to get the greyscale lib running on the COP?
09:01:53linuxstb_(s/CPU_INT/COP_INT/ in timer.c)
09:03:36NIXON-Trentlinuxstb_ i've got photo ipod..
09:03:37NIXON-Trentsorry was afk..
09:03:56 Quit ducbian ("Leaving")
09:04:36linuxstb_NIXON-Trent: Then you could try this - (I've never used it though, so can't answer any questions about it - and they would be off-topic here anyway)
09:05:17linuxstb_But if you do get it working, I would be curious to know how the playback quality compares to mpegplayer.
09:05:51 Join petur [0] (i=d4efd6a6@rockbox/developer/petur)
09:08:45 Join austriancoder [0] (n=austrian@
09:09:00austriancoderpetur: ping
09:09:06*petur waves to austriancoder
09:09:40austriancoderpetur: so.. i have looked at each stack and i must say every one has a minus point - check wiki
09:11:32austriancoderWASABI-h is fine, because it has all from ubsotg to classes, but it could be harder to port it to other usb chips
09:12:00linuxstb_austriancoder: Do any of them rely on a malloc/free implementation?
09:13:15 Quit GodEater ("CGI:IRC (Ping timeout)")
09:13:40NIXON-Trentyeah thats what i was asking about linuxstb_
09:13:52NIXON-Trentif it supported fastforward and rewind
09:14:03austriancoderlinuxstb_: WASABI-h... yes the others i am checking.. one moment...
09:14:14linuxstb_NIXON-Trent: It says it does on that page.
09:14:53linuxstb_austriancoder: That probably isn't a reason to reject a stack, but it's something you'll need to change.
09:15:05*petur is getting terrible work interrups :/
09:17:10peturlinuxstb_: I've been trying to port that wasabi stack to h300, the mallocs are used to allocate mem for the typical usb structures for each endpoint. That number should be limited so I created something with a fixed array of buffers
09:17:24peturnot optimal but just trying to get some life in it
09:17:45austriancoderpetur: so.. should we use the wasabi stack?
09:18:26peturI really don't know. What do you feel about the code structure?
09:18:31austriancoderLPC214x USB stack uses no free or malloc... usb4rt uses free and malloc
09:20:06linuxstb_Another question I would ask is if they all have active developer communities? It could be very helpful if there are devs around you could ask questions to.
09:20:45austriancoderpetur: the structre looks okay... but the interface to the hardware looks not so nice to me. Here i like the linux way much more.. but every stack has plus and minus points..
09:21:22peturI had e-mail contact with the author of wasabi some months ago, but he's not responding any more. He did like the fact I was using his code
09:22:27 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:22:28peturaustriancoder: yes, the hardware link is not so good, but then again it was only written for the isp
09:23:03 Join davina_ [0] (
09:23:58peturI'll try to look at that code too tonight.... damn real life and payed work
09:24:15austriancoderpetur: refactoring hardware interface and it should be okay... maybe we can take here some parts of usb4rt
09:24:36 Join B4gder [0] (
09:24:40peturmaybe.... have a look
09:25:41austriancoderelse I would start with a simple stack and use classes from wasabi
09:26:13 Quit Lynx_ (" HydraIRC -> <- The future of IRC")
09:26:21peturthat would be another option... but not much different from taking wasabi and changing the hardware interface
09:28:16peturalso, I have no clue how good wsabi is rdevice-related, I've only been poking around in the host related stuff
09:28:37petur(fix the typos yourself)
09:30:00austriancoderI have the same problem.. I have only looked at the source but I never seen it in action
09:33:24austriancoderpetur: Okay.. will use the wasabi stack for coding work. I will begin to import the stack in my local rockbox and I will spend some time to refactor the hardware interface.. is this a plan or not? *g*
09:34:23 Quit linuxstb_ ("CGI:IRC")
09:34:30austriancoderpetur: also you have done some work with it... so it would be no waste of time
09:34:37peturok... but we do want to use the hardware part for h300 ...
09:35:17peturI'll zip up what I made of it but it's a real mess. I only wanted to get it compile asap and get life in it, so the structure is a bit borked
09:35:35peturhmmm.... remove that 'bit' at the end
09:35:46peturand the 'a' too
09:36:47petursee if you can pinpoint the place where you can insert an interface to other hardware
09:38:04peturat some point, all usb drivers should be talking about endpoints and transfers, that would be the point to look for
09:39:24peturI've learned quite a bit about the isp1362, it's actually a nice chip. You give it a buffer and say: send this using that transfer type
09:39:47peturI don't know how other controllers do it, and certainly not the pp one
09:40:04austriancodersansas chip is a EHCI based one..
09:40:35 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:41:40peturI'll have to read up a bit on USB controller standards, don't know how EHCI looks like
09:41:44austriancoderwill be back in about 2 hours... need to do some shopping
09:41:57 Nick austriancoder is now known as ac_away (n=austrian@
09:42:07B4gderehci is quite similar to ohci ;-)
09:42:26B4gderbless you
09:43:11B4gderehci is a standard set of registers
09:43:26peturwhat an extremely short wikipedia article about them
09:45:39 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:45:44markunThe S3C2440 in the Gigabeat is OHCI compatible
09:49:30linuxstbIs that accessible via the dock?
09:51:46B4gderthere's 155 page pdf there
09:51:46 Join miepchen^schlaf [0] (
09:52:39LinusNbetter print it in two copies :-)
09:52:50printfXh4Did someone say... print!? :D
09:52:58peturbah... work stkov...
09:57:06markunLinusN: yes, it's is
09:57:14markuneh, that was for linuxstb
09:58:01 Join Lynx_ [0] (
09:59:44 Join Thundercloud [0] (n=thunderc@
10:02:27*linuxstb wonders if anyone has looked at
10:03:50*JdGordon looked at the text...
10:04:51LinusNlinuxstb: looks nice
10:05:23 Quit printfXh4 (Read error: 104 (Connection reset by peer))
10:05:32 Join printfXh4 [0] (
10:06:34JdGordonisnt ape the codec which is supposed t be x8-centric?
10:07:09Zagordionoea: you don't have to prepend commit comments with the filename. in fact I'd like to ask you not to.
10:07:21 Quit Thundercloud (Remote closed the connection)
10:07:54markunJdGordon: what do you mean?
10:08:19 Join Thundercloud [0] (n=thunderc@
10:08:55linuxstbJdGordon: Only so far as the bytestream consists of 32-bit little-endian integers (but the bytes are processed one at a time - so not in the same order they appear in the stream), and that the algorithm lends itself to MMX optimisations (vector math).
10:08:56JdGordonits heavily optimized for x86 and wasnt intended for DAPs? or am I tihking of a different one?
10:09:08linuxstbJdGordon: Yes, that's what's been written about it.
10:09:21JdGordonalrighty, then good work getting that far :)
10:10:33*JdGordon getting very sick of this list code ;p
10:13:12linuxstbJdGordon: Don't you have an exam? ;)
10:13:44JdGordonyeah, friday...
10:14:02JdGordonthen a weeks break, then 2 the week after... but im not worried :'(
10:14:30 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
10:17:51 Quit GodEater ("CGI:IRC")
10:18:16JdGordonare we allowd to use MAX() and MIN() macros? or best if we dont?
10:20:09 Join XavierGr [0] (
10:22:15linuxstbJdGordon: They're defined in system.h, and I can't think of a reason why you shouldn't use them.
10:22:31pondlifepetur: Commiserations. I've had work stkov for about 6 months now... Good luck in popping.
10:22:41 Join Rondom [0] (
10:23:22JdGordonlinuxstb: ah cool,
10:24:06peturpondlife: they just shifted priorities, and the board I'm working on is now top prio :(
10:25:33peturor as my boss told me friday: get prepared to do some extra time
10:25:48pondlifeextra time = extra pay?
10:26:02pondlifeNot so bad then.
10:26:10peturbut I'd rather have the extra time for myself
10:26:58*pondlife too
10:28:08peturB4gder: did you see that talk of Linus Torvalds about scm?
10:28:27peturhe made some famous quotes
10:28:36Zagorhe frequently does :)
10:29:05peturcvs users belong in a mental institution, etc...
10:29:21*linuxstb checks himself in
10:30:16peturif I can take my laptop and they have good internet access, I'm in
10:30:32JdGordoncvs ci self
10:30:38*JdGordon runs away
10:30:50ZagorLinus tends to get enthusiastic about his projects, which is generally a good thing. And he enjoys making pointed statements that fuel debate.
10:31:16ZagorHe'd be called a troll if in fact he wasn't the boss :-)
10:31:29printfXh4Did someone say... troll!? :D
10:32:58GaloisLinus trolling:
10:34:56 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
10:35:38ZagorGalois: but still you can't deny that he has a very valid point
10:36:37Zagorhe simply doesn't bother to try and be diplomatic
10:39:35 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
10:41:38GodEaterI found most of the arguments he made for using git very compelling, if a little blunt :)
10:42:07ZagorAbsolutely. In fact decentralized SCM like git would probably be a better option for rockbox than SVN. The problem is that it's a very difference concept and a very different tool, causing confusion and requiring training. So we chickened out and went for svn anyway.
10:42:22GodEaterI was thinking the same thing
10:42:41GodEateresp. his points around "not checking in new code to cvs until it passes the test suite"
10:42:53GodEaterclearly we don't have a test suite - but we do like patches that work :)
10:43:06GodEaterif git were being used then patches wouldn't need to be in flyspray
10:43:18GodEaterthey could exist in developer X's repo for the interested party
10:43:29GodEaterand only get merged into the "master" repo when everyone liked them
10:43:42Zagora problem with that is that many people do not have machines where they can publish their repos.
10:43:48JdGordonso yo end up withh hundreds of forks/
10:43:51GodEatera salient point
10:44:07GodEaterJdGordon: yes, but git does make merging branches very easy
10:44:11ZagorJdGordon: yes, one fork per patch. are patches in text files really better? hosts git repositories for hundreds of kernel developers
10:44:48GodEaterlinus said he does merges across 22,000 files 4 or 5 times a day
10:44:52ZagorGalois: yeah, we'd need something like that too
10:45:00GodEaterin less than 5 seconds or so per merge
10:45:22GodEaterZagor: that's a lot of disk space!
10:45:30Zagordisk is cheap
10:45:39GodEateralthough granted we don't have *hundreds* of developers
10:45:58 Join Nico_P [0] (
10:46:28 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
10:46:51GodEaterI also liked what he said about the "commit access" politics
10:47:04JdGordoncould git and svn be run in parallel?
10:47:23 Join Rob2222 [0] (
10:47:23ZagorJdGordon: sort of, but with more work than gain imho
10:47:35GodEateryou could maintain a local git repository of what's in svn
10:47:37***Saving seen data "./dancer.seen"
10:47:42GodEaterand use git for your local work
10:47:46JdGordonGodEater: I think he got that wrong, you have worse politics, because people say fine dont take my patch ill fork and no contribute to you again
10:48:14GodEaterJdGordon: that's a possibility certainly - he didn't say it got *rid* of politics :)
10:48:19ZagorJdGordon: generally that's not a problem for larger projects. developers want their code in the main branch.
10:48:39GodEaterbut we have a similar "problem" with flyspray
10:48:52GodEaterpeople have patches in their they'd like to see commited
10:49:08GodEaterand we don't all agree they should be in there
10:49:13Zagoraccess politics is a natural consequence of a centralized repository. of course anyone can commit when "commit" is local.
10:49:15GodEaterI don't see that moving to git would change that
10:49:16JdGordonyeah, but they dont have a dozen patches in one task, which effctivly would happen with git
10:49:36GodEaterJdGordon: that would mean they weren't using git correctly
10:49:51GodEaterLinus's ideal is that you have one branch per "feature" or patch
10:49:53JdGordonah, ok
10:50:24JdGordonI dont think he expained that well then, or I tuned out during that part :p
10:50:25GodEaterso part's of a developer's repo could be commited, whilst leaving behind something else that other's don't like as much
10:50:34GodEaterit *was* a long talk :)
10:50:51GodEaterI probably missed bits too by trying to read my email at the same time
10:50:55JdGordonwas still better than studing though
10:51:03GodEaterhahaha - what isn't ? :)
10:52:51GodEaterI think the mindset for a developer keeping *all* their changes in one branch is as a result of the difficulty you get with merging in CVS. When it becomes as easy as it is with git, it's more natural to create one branch per patch
10:53:08JdGordondinner :) bbl
10:53:10NIXON-Trenteek just tried to install a boot loader for ipod
10:53:16NIXON-Trentnow ipod displays a folder with ! on it
10:53:19NIXON-Trentand the ipod url ...
10:53:26GodEaterI'm not really in a position to comment though, I've not done a lot of work with either system
10:53:49ZagorGodEater: not entirely. it's also the feeling of making accumulating improvements
10:54:37GodEaterZagor: that's natural when the features are in a working state - but there'd be no point accumulating multiple "work in progress" features
10:54:47GodEaterIMHO anyway
10:55:47NIXON-Trentanyone know a simple way to get linuxipod + rockbox + apple os on a boot loader screen ?
10:56:06linuxstbCould SVN branches be set up for major upcoming features? I'm thinking the GSoC projects as an example...
10:56:07GodEateripodlinux is offtopic here
10:56:24NIXON-Trentwell anyone know how to get a boot loader
10:56:29NIXON-Trentthat shows rockbox + apple
10:56:33NIXON-Trentand i can pick which to load ?
10:56:42linuxstbIf you installed Rockbox according to the installation instructions, you have that.
10:56:46GodEateripl's bootloader, which is what you're talking about, is also offtopic
10:56:54Zagorlinuxstb: I've been thinking about a system to promote more branches. But the access system is the biggest stumble.
10:57:06*GodEater votes we move to git :)
10:57:17*GodEater is fully aware his vote is worth nothing.
10:57:21NIXON-Trentlol whats not off topic lmao
10:57:25 Join maffe [0] (n=maffe@
10:57:35NIXON-Trentlinuxstb what do u mean if i installed rockbox ?
10:57:37scorcheNIXON-Trent: rockbox related topics, unsuprisingly
10:57:48linuxstbZagor: For SoC, 2 of the students already have commit access, and I can't see a problem with giving the other two students access - with "social" restrictions.
10:58:37linuxstbNIXON-Trent: The Rockbox bootloader you installed lets you boot either Rockbox or the Apple firmware. You just need to make the decision before turning on your ipod (and then turn the hold switch on or not)
10:58:40*GodEater looks forward to hearing what these social restrictions are
10:58:52 Quit Kohlrabi (Remote closed the connection)
10:58:54NIXON-Trentlinux any alternatives to that
10:58:56GodEaterare we not allowed to talk tot hem ? :)
10:58:59Zagoragreed. my idea was a more general concept, to get some of the git benefits (branches instead of patches)
10:59:01NIXON-Trentincase i want to choose from 3 boot options ?
10:59:23GodEaterNIXON-Trent: the rockbox bootloader will also boot iPL
10:59:31GodEateryou hold play when booting
10:59:54linuxstbNIXON-Trent: If you install IPL according to their instructions, you'll install their bootloader, which will also detect Rockbox and give you an option to load that.
11:00:04NIXON-TrentIPL ?
11:00:31NIXON-Trenti was trying to install that
11:00:39NIXON-Trentand well it seemed to bugger my ipod and i just had to restore it then :/
11:00:53NIXON-Trent<3's to rockbox being way easyer :P
11:01:08NIXON-Trentonly reason i want ipodlinux is because u can rewind and fast forward movies :/
11:01:46 Quit GodEater ("CGI:IRC")
11:01:59 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
11:03:09 Quit Rob222241 (Read error: 110 (Connection timed out))
11:03:46 Quit lids (Remote closed the connection)
11:04:19 Join Kohlrabi [0] (
11:04:57 Quit jhMikeS (Nick collision from services.)
11:05:03 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
11:05:14 Join lids [0] (i=lds@gateway/tor/x-abad39361930f3b0)
11:06:57 Quit idnar (Nick collision from services.)
11:06:59 Join idnar_ [0] (n=mithrand@unaffiliated/idnar)
11:11:12 Quit aliask ("ChatZilla [Firefox]")
11:11:23 Quit SaLoMoN (Read error: 110 (Connection timed out))
11:12:35 Join SaLoMoN [0] (
11:19:00 Join obo [0] (
11:20:09 Join desowin [0] (
11:21:00 Quit ze (Read error: 104 (Connection reset by peer))
11:21:12 Join ze [0] (
11:23:56 Quit desowin (Client Quit)
11:24:28 Join spiorf [0] (n=spiorf@
11:28:16JdGordonanyone interested in testing out my new list code which I'm pretty sure fixes all known bugs because of this silly accelleration stuff?
11:29:18linuxstbWere you able to reproduce the Sansa problem with slow scrolling?
11:30:03JdGordonyeah, the problem is the button dirver doesnt send button_rel events which is confusing te action code
11:30:42linuxstbDid you manage to fix that, or is your patch for other bugs?
11:30:51JdGordonthe easiest fix is to check how long since the last action, but havnt got that working yet
11:31:20LinusNso what is your current fix then?
11:32:25 Nick ac_away is now known as austriancoder (n=austrian@
11:33:15JdGordonFS #7258
11:33:20 Join Rick| [0] (n=rhelmus@
11:33:45JdGordonmakes the selection code more similar to how it used to be, moving up/down the list works nicely again
11:33:46 Join ljcrabs2 [0] (
11:35:50LinusNok, so this doesn't really solve the acceleration problems, but merely the issues with the list cursor?
11:36:12Rick|midkay: In case you are here; I wanted to thank you because my iPod is now working again :)
11:36:18LinusNi see
11:36:51JdGordonwhich accell problems? I thought the main problem was the bad language for the settings? other than that it works doesnt ti (except that wierd sansa bug)
11:38:31austriancoderpetur: I am waiting for your mail
11:38:32 Join Entasis [0] (
11:38:45linuxstbJdGordon: I think the problem is on the wheel targets - the acceleration needs to take account of the speed, but it doesn't. I like how it works on button targets though.
11:39:26peturaustriancoder: it will be tonight, don't have it with me.... and lots of work here :/
11:39:35pondlifeJdGordon: Should this patch also fix the Iriver LCD remote problem?
11:39:41austriancoderpetur: no problem
11:40:26JdGordonpondlife: hopefully, i've been testing in the h300 sim,and found no problems
11:40:34NIXON-Trentgod ipodlinux is so hard to install things on
11:40:48pondlifeJdGordon: Urgh, something is horrid with SVN+the patch though...
11:41:13pondlifeIt's hard to describe but I'll try and put a note on Flyspray
11:41:49GodEaterNIXON-Trent: again, iPL is not what we want to discuss here. Please take it to #ipodlinux.
11:41:50JdGordonalso, I tried reporoducing your follow playlist bug (before these changes) and couldnt
11:42:12pondlifeI don't think that's list related - it's buffering related
11:42:28pondlifeWith your patch it doesn't allow the cursor bar to go to the top or bottom of the screen before scrolling.
11:43:38pondlifeI'll try and explain with pictures on Flyspray...
11:50:35pondlifeJdGordon: OK, look at
11:52:48pondlifeLet me know if that makes sense.
11:52:58JdGordonit does with the pics :)
11:53:02JdGordonbut I thought I fixed that
11:53:18pondlifeCan happen with both UP and DOWN.
11:53:50JdGordondefiantly works correctly here...
11:54:10pondlifeDo you want my config?
11:54:23JdGordonoh wait... hmm..
11:54:30JdGordonit happens when you swap directions
11:54:52pondlifeAnd there are undisplayed items in the direction of travel.
11:55:51JdGordonthis one is going to be fun to fix :'(
11:56:26pondlifeIs it me, or has the list code got over-complex recently?
11:56:27pondlife ;)
11:56:38 Quit Bagder (Read error: 110 (Connection timed out))
11:56:43pondlifeCan it not be KISSed back to health?
11:56:45JdGordonno, its always been bad
11:56:50pondlifeAh, ok
11:56:54JdGordonthats what this was doing...
11:57:03pondlifeGood, KISS it better then!!
11:58:30JdGordondoes anyone else think it would be a good/(fun?) idea to start the apps/ from scratch?
11:58:33scorcheZagor: i am only 20-min into the talk, but how does distribution work with telling people to download "our binary"?....if it is answered later, feel free to tell me "WTFV"
11:58:57pondlifeJdGordon: Yep, time for Rockbox NT :/
11:59:31JdGordonscorche: the idea is that people grab whichever they want...
12:00:13scorchesounds like a headache to support
12:00:33 Part maffe
12:00:40JdGordonpondlife: im semi serious... start apps from scratch so we dont have to keep dealin with 3+ year old code....
12:00:43 Join maffe [0] (n=maffe@
12:00:45pixelmalinuxstb: just wanted to add a correction to your post - but then I saw you already edited it ;)
12:01:00LinusNJdGordon: 3 year old *debugged* code
12:01:05JdGordonwell yeah
12:01:24LinusNbut i understand the feeling
12:01:27ljcrabs2On the ipod with a dock connected hardware, is it the ipod software or hardware that handles commands, eg mute/next track etc?
12:01:31LinusNrefactoring isn't all bad
12:01:40JdGordonbut its 3 yr old code + multiple hacks for each feature...
12:01:51GodEaterscorche: "our binary" would be one build from the "master" repo at
12:02:07GodEatermuch as the "master" linux kernel tree is kept at
12:02:37LinusNJdGordon: we could/should rewrite large part of rockbox, and we often do
12:02:38 Join rift [0] (
12:02:41JdGordonCould we possibly start a experimental branch for this?
12:02:44scorcheGodEater: and this "master" would still have to have multiple guardians to say "we want this part and not this", no?
12:02:53linuxstbpixelma: Yes, I'm always discovering new things Rockbox can do...
12:02:54LinusNlike the action concept, or the settings code
12:03:09LinusNtwo examples of recent refactoring
12:03:10GodEaterscorche: yes it would
12:03:13JdGordonLinusN: but that even causes other problems, refactoring still depends on other hacks which we just cant get rid of anymore
12:03:42LinusNJdGordon: you should read "The UNIX Philosophy"
12:03:43GodEaterscorche: well - in theory it could be one - but in practise I suspect we'd still end up with our "core" devs deciding amongst themselves what to pull in
12:04:56LinusNwe won't be able to rewrite rockbox without buildind a monstrous "second system of man"
12:05:29scorchehmmm...i could see where it would simplify branches and big changes, but other things, it would just add extra responsibilities onto people, and small changes would be simpler just with posting patches
12:05:56 Quit B4gder ("It is time to say MOOO")
12:05:57*scorche resolves to shut up and make comments after viewing the entire vid
12:06:05*LinusN goes to lunch
12:06:39GodEaterscorche: the benefits of git would be felt most by the devs themselves, I accept that it would possibly make life more complicated in the short term for the master build system maintainer
12:06:58dionoealinuxstb: are you planing on doing "rockbox independant" releases of your monkey audio decoding lib ?
12:07:13linuxstbdionoea: Probably, yes.
12:07:28scorcheGodEater: i realize that, but...hmmm....i will talk to you about it in 40 min
12:07:30dionoeacool :)
12:08:06GodEaterscorche: heh - no problem. It's very academic at the moment given the change to svn only 6 months ago :)
12:08:33JdGordonwouldnt that make it a better time to convert then later on when we are even more entrenched in svn?
12:09:19linuxstbdionoea: Although assuming no-one objects, I would like to use Rockbox SVN as the "master" repository for it, so I don't have to worry about maintaining multiple versions. But I need to think of a way to unobtrusively add the standalone parts to Rockbox's SVN (it's only a few small files).
12:10:27dionoealinuxstb: I'm sure that once media players begin using it you'll get ASM patches to optimize it. I don't know if that would be ok in the rockbox source (It would be ifdefed out easily)
12:11:57linuxstbdionoea: I still need to decide on a final license for it. Currently it's GPL'd (mainly because I've taken some code from another GPL project), but I'm aware that may cause problems for other projects.
12:13:08dionoeait wouldn't for VLC :)
12:13:28peturI don't know about others, but in the end you're already working loaclly in a git way, having multiple checked out working copies to do multiple things
12:13:51scorcheit does seem like it would simplify many things
12:14:23linuxstbdionoea: ffmpeg is the main project I'm thinking about. It would be nice to get the benefit of the skills of their developers.
12:14:41dionoeaThey have GPL and LGPL parts ... but they don't like external libs.
12:14:42peturbut you would need a maintainer that pulls changes into the master, unlike now where every dev with commit access pushes it into the repo
12:14:52dionoeaThey always try to write their own codecs
12:14:52 Quit lids (Remote closed the connection)
12:15:05scorcheand make collaborating much easier as well, but there are still a few things i am leary about (which i can address after viewing this in its entirety) seems to put a lot more load on the maintainer(s) of the head
12:15:51GodEaterscorche: yes, it's the one thing I didn't spot in the talk - how that part is managed.
12:16:00linuxstbdionoea: Do you know about any other open source media players? Would they be happy with a GPL'd lib?
12:16:09GodEaterscorche: although the kernel doesn't suffer from having to distrubute binaries like we do
12:16:21GodEaterscorche: which makes life a hell of a lot easier
12:16:31dionoealinuxstb: MPlayer and Xine would.
12:16:42linuxstbdionoea: My feeling is that proprietory software can just use the official decoder (which I assume they all do already). It's mainly gpl'd software that has problems with Monkey's.
12:18:26GodEaterscorche: I think the issue here really is that Linus is viewed as the "god" of kernel development. He maintains his own repo, and that's what most people trust. With RB we don't have someone who's the "father of dev", so we don't have a key repo to trust for use as the "master".
12:18:31scorcheoverall, at this moment, my opinions are: For collaboration between devs and for experimental branches/projects, it is great and a huge improvement; It comes at a huge cost of putting a ton of load on the maintainer(s)
12:18:44 Quit GodEater ("CGI:IRC")
12:18:57 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
12:18:57scorcheGodEater: and he is perfectly willing and has tons of time to be that master
12:19:11GodEaterscorche: well, he gets paid for it :)
12:20:09linuxstbI don't know if this is because we don't have branches in the code, but I've noticed that it's quite rare for Rockbox devs to collaborate on new features. It's mainly people working alone - which isn't always the most productive. So a more flexible scm could help that.
12:20:27scorchefor git, it expends more time to say "hey <maintainer>, check this out please" "alright...hold on" <other discussions> "cool...i am pulling lines xxx in...should i pull xxxx as well?"
12:21:22scorchebasically, it seems like it comes down to "Is the flexibility and all that really worth all this extra load that a few get to carry?"
12:22:10scorcheas a developer, it seems obvious which choice to go with, but if one is to think of others...
12:22:15JdGordonunless git allows a group of maitainers on the master, so we would still push changes instead of having them pulled like linus keeps saying
12:23:09scorcheim sure it does (at least i hope so), but it still adds some extra effort for some
12:23:22LloreanI think the real question is, do we currently have a problem that will be solved by better SCM?
12:23:48LloreanWill it get some of these patches into the repository more quickly than currently, and will it allow the better sort of collaboration linuxstb was talking about or will we still have the same sort of lone-wolf working going on?
12:23:58scorchewell, i wouldnt say it is a problem, but git simplifies many things and allows much more room for better collaboration
12:24:28scorcheat least it seems to, at this point
12:24:53LloreanBut is SVN getting in the way of collaboration currently? It seems like most people currently tend toward working on a feature themselves, and the question is if that's because it's hard to collaborate, or just how the people around here tend to do things?
12:25:03GodEaterI think for the sort of work we see our core devs do, where they submit patches to flyspray, git makes more sense
12:25:05*scorche wonders if we can pull linusT in here and have a little talk
12:25:20*JdGordon suspects its more than just svn that stops the sort of collaboration that would be nice...
12:25:55scorcheSVN does for a is much harder and more effort to have to apply patches and such when working on each project
12:28:17scorcheit makes things much simpler and faster to ignore the patches and just pull things here and there
12:28:26peturI feel git is better for big projects, and rockbox isn't that big so the benefits would not be much
12:29:09 Join arkascha [0] (
12:29:21scorchein terms of the benefit with collaboration, i feel it is, but as for the duty of the maintainers, i am not sure if it is big enough as well
12:29:30 Part arkascha ("Konversation terminated!")
12:29:34peturscorche: you would still need a place where people can advertize their 'branches' and others to comment on. I think I like flyspray much
12:31:14GodEaterpetur: this is very true as well
12:31:20scorchei like it a lot as well, but the whole "here is a patch of some part of xxxxxunfinishedwork" can be improved
12:31:41LloreanI think one of the major problems of multiple GIT branches will be that people will probably just use them for unsupported builds.
12:31:45linuxstbAlso, would git retain the history of the patch when it was merged?
12:31:55LloreanYou'll have 50 branches that all have 20 different modifications, all nearly identical
12:32:10LloreanUnless it's possible to easily prevent that.
12:32:25*petur says: lunch!
12:32:28 Join norbusan [0] (
12:32:36LloreanThere will be a lot of non-development cruft trying to float in, I think, unless it's easy to actively prevent this.
12:33:01scorcheLlorean: yes, but i am mainly thinking about current committers and people who write patches often
12:33:04JdGordonLlorean: yeah, thats what I thought also, but you'd have an official set of branches, and then the rest hostig their own
12:33:15linuxstbLlorean: I think the intention would be one branch for each patch. So if anyone tried to use them for unofficial builds, they would be breaking that rule.
12:33:47scorcheLlorean: we would be relying on a head branch for binaries and such
12:33:51Lloreanlinuxstb: Yeah, then someone has to actively keep track of the rule-breaking ones and do whatever's necessary to remove them.
12:33:57JdGordonyou'd still have the problem where patches/branches fall out of sync with the master wouldnt you? git doesnt magixcally fix that?
12:34:03 Part norbusan
12:34:08scorcheLlorean: not really, if i am understanding this right
12:34:14linuxstbLlorean: That's the same as flyspray's "one feature per task" rule.
12:34:41LloreanAnd that still doesn't always get enforced.
12:34:49scorcheJdGordon: yes, but it would be much easier to re-sync things, i would think
12:35:01JdGordonusually only on tassks which are not going t be commited anyway, i.e AA
12:35:23LloreanWhat happens when person A abandons a task, and person B picks it up?
12:35:30LloreanWith Git do both have separate repositories?
12:35:34JdGordonscorche: how could it? if 1 line changes, the same line would have to be fixed by a maintainr regarless of the scm
12:35:41GodEaterLlorean: yes they would
12:35:46LloreanGodEater: I don't think I like that
12:35:51GodEatereveryone has their own copy of the reop
12:35:54LloreanWith Flyspray you just post it on the end of the existing task.
12:36:05GodEaterso ?
12:36:13LloreanSo if Patch X gets worked on by seven people, you can see where it ends up.
12:36:28JdGordonyep, massivly massive amonuts of forks
12:36:29GodEaterLlorean: but person B has pulled in person A's work
12:36:39LloreanGodEater: Assuming you know about person B.
12:36:39GodEaterLlorean: complete with history
12:36:54LloreanAnd assuming Person C has noticed and hasn't done some of his own work, with his own fork of the task.
12:36:55scorcheJdGordon: because you dont have to work with patches and pull it from their source itself
12:37:04GodEaterwell if we're still talking about using flyspray, instead of posting a patch
12:37:08GodEateryou post a link to your repo
12:37:09LloreanThe presentation of Flyspray is more linear, which I think just makes it clearer to anyone picking up that it's one task.
12:37:19 Quit jhMikeS (Nick collision from services.)
12:37:21GodEaterI fail to see how this would stop people using Flyspray
12:37:23LloreanMultiple repositories I think could lead to someone missing that person B took over, and restarting from Person A's work
12:37:25 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
12:37:37GodEateryou're just not posting files anymore
12:37:43LloreanWell then what's the real benefit?
12:37:50scorcheLlorean: which is why this is all about collaboration which includes communication
12:37:53GodEaterthe benefit is to the DEVELOPER
12:38:05LloreanNot "to whom is the benefit"
12:38:09Llorean"What is the real benefit"
12:38:50JdGordondoes anyone know how to get a diff of the whole tree btween two revisions? (in svn of couse :p )
12:39:32LloreanGodEater: I understand pulling to the main repository is easier than all this work with patches.
12:39:44GodEaterLlorean: the benefit I see is to the dev. let me explain :
12:39:57GodEaterLet's say your developer A, you're working on feature X.
12:40:12GodEaterdeveloper B posts a patch to supprot feature Y, and wants you to test it.
12:40:15*scorche deletes his response to see where this is going
12:40:29GodEaterwith svn, you've got only your one repository to manage easily
12:40:45GodEaterwhich already contains "polluted" code from your feature A.
12:40:54GodEateryou dont' really want to add more stuff at this time
12:41:17GodEaterso you'd have to check out a whole new repo, and patch in feature Y to make sure it doesn't mess with your work on feature X
12:41:30scorcheLlorean: honestly, i am not really confident in my ability to explain it properly...i would recommend watching the video of the talk
12:41:31GodEaterin order to statisfy developer B's request for testing
12:41:52GodEaterwith git, you've already taken out the full "master" repo
12:41:57GodEateryouve' branched it for feature X
12:42:21GodEaterand now to satisfy developer B's request, you simply branch again off of "master" to include feature Y
12:42:29GodEaterthis hasn't messed with your original work at all
12:42:34GodEaterand you've still only got one repo
12:42:45GodEaternow you find feature Y is working great.
12:42:55GodEateryou merge it back into your copy of master, AND your feature X branch
12:42:58GodEaterand continue working
12:43:06GodEaterall with a minimum off fuss
12:43:15LloreanSee, I just have a small script that takes a diff of my current sources, saves it, and reverts, and another that reverts and reapplies that diff, making it easy to test an individual patch and restore from the 'bookmark' with a minimum of fuss.
12:43:53GodEaterLlorean: but that won't let you merge both features in until they both get commited to the master tree
12:44:10GodEaterand you check them out one revision at a time
12:44:37LloreanI'm not so sure why I'd want to keep the feature around unless it was going to be committed anyway.
12:44:45GodEatermy case there is extremely over simplified too - it only deals with two patches at once
12:44:46JdGordondoes anyone know what to do about a lang string which I want to deperacte but expect to reuse it later?
12:44:59scorchehmmm...he describes a bit about using both SVN and git at the same time in somewhere around 37 mins remaining
12:45:10GodEaterLlorean: perhaps you'd like to WORK on the feature? :)
12:45:20LloreanGodEater: Then I reapply my diff without reverting first.
12:45:31GodEaterscorche: that's the unofficial use of git in organisations bit I assume ? :)
12:45:41JdGordonLlorean: see, thats the svn way of thinking...
12:45:54GodEaterLlorean: yes, but then two days in you think, actually, no I didn't want it after all. No you're in a mess.
12:46:25LloreanGodEater: And if their diff won't remove cleanly, why would the git branch remove cleanly?
12:46:31 Join Kanashimi [0] (
12:46:36GodEaterit's not being removed
12:46:46GodEateryou've got it all on it's own in it's own branch
12:46:58linuxstbJdGordon: If you want to reuse it later, then it's not deprecated...
12:47:00GodEaterif you've done it right
12:47:15LloreanGodEater: Ah, I thought you were talking about incorporating it into the code you're currently working on for some reason.
12:47:29JdGordonlinuxstb: So just change the string to ""? so it doesnt waste bytes?
12:47:33GodEaterLlorean: you might have done, but you've branched the code you were working on first :)
12:47:39***Saving seen data "./dancer.seen"
12:47:41LloreanGodEater: All it sounds like is a way to manage .diffs
12:47:52GodEaterLlorean: that's all any SCM is :)
12:47:55linuxstbJdGordon: I would just keep it as it is.
12:48:05LloreanBut it sounds like a way to manage diffs that makes other things more complicated
12:48:16 Join DerPapst [0] (
12:48:17GodEaterLlorean: I don't pretend to be an expert by any means, these are just the advantages *I* see.
12:48:19LloreanRather than simply having a folder of diffs and applying the one you want to work on before you start, and taking a diff when you're done working
12:48:46GodEaterperhaps to you - I don't find it that complex, and I personally don't like your chosen solution.
12:48:53GodEaterbut we're allowed to disagree :)
12:48:57 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
12:49:09scorcheLlorean: (really, watch the vid....i am trying to not comment much until i do, so as to understand it more)
12:49:19GodEateryeah definitely watch the video.
12:49:26safetydangit would be good for long running development patches like the lang cleanup
12:49:26DerPapstJdGordon: in case you haven't looked it up already: svn -r diff 13337:13456 > blah.diff
12:49:28GodEaterthe google people he's talking to ask smart question :)
12:49:39JdGordonDerPapst: thanks, found it :)
12:50:11 Join lids [0] (i=lds@gateway/tor/x-34524e49d08945d8)
12:50:15KanashimiHello there. I hope no one minds me asking a non-rockbox specific question here. Does anyone here happen to own a Cowon A2 player? I've been hearing it has trouble with MP3 files encoded with LAME -new VBR. I ordered mine a bit ago and just found out about the issues, so I'm a bit worried. How bad is the glitchyness? I don't necessarily need perfect audio for the moment,
12:50:20Kanashimibut hopefully it's not too terrible?
12:50:35scorcheKanashimi: we do mind =)
12:50:59*GodEater runs to get a sandwich
12:51:18linuxstbKanashimi: Your question is off-topic, but I also don't think anyone here will know the answer - the A2 isn't supported by Rockbox, so owners of it don't hang out here.
12:51:45 Quit Brunellus (Read error: 110 (Connection timed out))
12:52:51KanashimiAlright, sorry about the bother :) Been using Rockbox on a iRiver of mine for a while and have visited before too. Just kinda worried and was hoping I might find someone by chance.
12:54:30 Join norbusan [0] (
12:56:13 Part norbusan
12:56:19Galoisbesides, it supports ogg, so why bother with mp3
12:57:08Galoisunless you happen to have some mp3 files for which you do not have a lossless copy (hmm, wonder how that could happen)
12:57:59DerPapstmaybe some users don't like ogg?
12:58:17markunpretty pointless discussion
12:58:18LloreanThere's also online music stores, that aren't even in gray areas, that sell MP3s these days
12:58:39linuxstbOr the CD being lost/damaged/stolen, ...
12:59:02markunlinuxstb: then you have to destroy the MP3 files as well of course :)
12:59:02Galoiswell the vast majority of legitimate online sales are in AAC format, and I don't see Cowon users complaining about lack of AAC support
12:59:28linuxstbGalois: So all MP3 users are thieves? ;)
12:59:40markunGalois: what kind of complaints do they have?
13:00:01LloreanGalois: Don't forget they're DRM wrapped AAC, so you can't legally use them in the US even on an AAC supporting player
13:00:47GaloisI mean, even the non-DRM files offered for purchase are in AAC format
13:00:55KanashimiHeh, let's not start on that kind of tangents. Well, there are a couple of things. While I do like ogg as a format it does also seem to require more processing which reduces battery life. And it is after all still the most common music format out there, used by independent artists and what not.
13:01:45safetydanAnyone looked at git-svn? Might be a nice halfway house for those who want git without forcing everyone to use it.
13:01:46Galoisdoes −−preset invoke −−vbr-new?
13:01:50LloreanGalois: The majority are DRM wrapped AAC. Then DRM Wrapped WMA from the Playsforsure stores, and MP3 in most of the non-DRM stores I've seen. Which store did you have in mind that was both a majority of online music sales, and DRM-free AAC?
13:02:26*DerPapst just got a comercial for an USB turntable ^^
13:02:42Galoissorry, I wasn't very clear: all files sold by the leading vendor of legal online music sales, whether DRM'd or not, are in AAC format
13:02:58obosafetydan: did a pull last night - setup a few branches, seemed to work okay
13:03:00LloreanBut until you can support the DRM, lacking AAC support doesn't matter
13:03:07LloreanSo expecting complaints about it is silly
13:03:15markunGalois: while looking for music under the creative commons, all music came in mp3 and not all of them provided the music lossless as well
13:03:25scorchesafetydan: i was thinking something like how he said it in the answer to a question, but i really think it would be better discussed when people have done more research on all this, so we can speak from a level platform
13:03:27linuxstbobo, safetydan: What would "git-svn" provide the typical Rockbox dev?
13:03:44safetydanlinuxstb: local branches
13:03:58GodEaterlinuxstb: a local git repository, with all git's features, that's easy to keep in sync with svn
13:04:08GodEaterlinuxstb: provided you keep proper track of your own branches :)
13:04:21GodEaterI believe it's what dana uses
13:04:32markunDerPapst: I've seen a guy DJing at a party with a USB turntable, worked very well
13:04:32safetydanscorche: true... I've just been looking at SCM systems for work related purposes lately as well
13:04:56linuxstbWould someone like to write a wiki page describing how git-svn could be used with Rockbox?
13:05:08linuxstbSounds useful to me - I would source control in my local branches...
13:05:42GodEaterdan_a perhaps rather.
13:05:50scorchei wouldnt mind, provided i had a bit of help (after i get done researching and viewing all this)
13:06:05JdGordon <- me skimming...
13:06:08GodEatergit also provides other more esoteric features
13:06:10safetydanWell I've only just typed "apt-get install git-svn" so...
13:06:15*scorche volunteers GodEater to help him
13:06:17LloreanDoes git-svn still focus on the mentality that there's one key branch, and nobody really gets to commit to it, instead it has to pull?
13:06:18DerPapstmarkun: i've never seen such a thing... but it's cool though. i mean there are washing machines that can use internet.. how usefull is that?!
13:06:24GodEaterscorche: happy to
13:06:55*scorche wonders when Llorean will have a chance to watch the vid =P
13:06:57GodEatergit-svn just pulls a central svn repo into a local git repo
13:07:05Lloreanscorche: I'm watching it.
13:07:13safetydanLlorean: git-svn has no impact outside of an individual developers machine. It just means I can track my change sets locally before pushing it to svn.
13:07:36oboalthough git-svn can also commit back to SVN
13:07:41GodEaterthis is true
13:07:45Lloreansafetydan: Ah, so it's still a commit based system, but locally you're effectively using git?
13:07:50safetydanLlorean: yup
13:09:14GodEatergit-svn would mean no changes necessary at the rockbox server end - it still just runs svn
13:09:55markunDerPapst: don't know if we were talking about the same thing:
13:10:16scorchegit-svn also seems to be a way to lessen the load on what would be put on the head maintainer(s) under using pure git
13:10:37GodEatergit-svn still requires commit access
13:10:47safetydanLlorean: though because you're using git, it's possible for developers to share git repos while developing patches and still keep track of who did what.
13:11:24scorchei do like how git tracks "content" instead of "individual files"
13:11:28Lloreansafetydan: That sounds better to me.
13:12:06GodEateralthough the keeping track of who did what is only maintained in those dev's git repo's
13:12:19DerPapstmarkun: i meant something like this:
13:12:20GodEateras far as the central repo is concerned, it was all done by whoever finally commits it
13:12:21LloreanI don't really like getting rid of the idea of commit access, because then you're depending on the central maintainer to do all the work of pulling.
13:12:36scorcheGodEater: yeah...that would be one strong limitation of it
13:12:47GodEaterLlorean: it's still automatable - it would just take work
13:12:53LloreanIt makes sense when you're doing a lot of non-live work, among a lot of people.
13:13:15LloreanBut with Rockbox we're committing code multiple times per day, from many directions.
13:13:25GodEaterstill doable
13:13:47GodEaterI envisage a script that connects to each dev's git repo
13:13:53LloreanTechnically speaking, anything is doable.
13:13:58GodEaterand reads a file containing a list of branches to pull
13:14:21GodEaterall the dev has to do is put a branch in there for it to get pulled to the master repo
13:14:32GodEaterthe script checks say every 15 minutes ?
13:14:55GodEaterbit rough and ready, but it would work
13:14:58scorchelinus actually talks about that in the vid too...he mentions a part about when there are issues with the merge....but, yes, atm in my research, it seems like it puts more load on the maintainer(s) which i dont like...i need to research a bit more to figure out exactly the amount we are talking here, though
13:15:05LloreanThis depends on single development machines, or single internet access points (or domain names), etc.
13:15:21obowhy not git-push?
13:15:40*scorche adds git-push to the list of things to check out
13:15:50*GodEater hasn't come across git-push yet either
13:15:59obothose who already have svn commit access could push to a central repo
13:16:02Lloreangit-push brings us back, mostly, to why not just svn-git?
13:16:07LloreanOr was it git-svn?
13:16:49amiconndionoea: Do you remember what the problem was regarding xobox on archos?
13:16:54GodEaterI don't understand the context of a "ref" with git-push
13:17:07amiconnIf it was speed, could it be that your recent changes make it run fast enough?
13:17:43 Join Arathis [0] (
13:17:46JdGordonI assume git-push is the equivilant of git-svn excecpt the mster would be running git instead of svn?
13:17:58GodEaterJdGordon: that's not clear from it's man page
13:18:16GodEaterat least, not to me :)
13:18:38*GodEater accepts he may be both stupid and ugly as definted by LinusT
13:19:03scorcheoye...according to him, if you disagree with him, you are ugly and stupid
13:19:17GodEaterexcept if you work on the KDE project :)
13:19:20LloreanGodEater, Scorche: Just to clarify, I do see the advantages of git in a lot of situations, I'm just not sure Rockbox is one of them.
13:19:25JdGordonI honestly cant see how having 1 maintainer could possibly work for any project, that would be a full time job and a bloody nighmare
13:19:45 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
13:19:50scorcheJdGordon: i was mentioning that before, but as i said, i dont really knwo how much work this would be
13:19:51GodEaterJdGordon: I'm actually considering emailing him and asking him about rockbox's situation :)
13:20:04LloreanJdGordon: In software that does actual release versions, someone or a team of someones has to define what is a release, and you have programmers with objectives. In that situation it makes a lot more sense.
13:20:13scorcheGodEater: same, although it would be easier to just ask him to come in ehre ;)
13:20:18GodEaterscorche: if all the devs went with git-svn, there's be no extra work :)
13:20:30GodEaterbut we wouldn't get all the benefits either
13:20:45scorcheGodEater: it has its well as performance issues comparatively
13:21:05GodEaterI've not used it yet, so couldn't comment
13:21:13GodEaterI've only used git on it's own
13:21:19JdGordonhmm... seems git can do a push anyway
13:21:36JdGordonalthough it seems more cmplicated than svn commit
13:21:42JdGordonGodEater: go for it :)
13:22:44scorcheJdGordon: well, actually, he is saying most things in his talk about a project that makes rockbox look miniscule, but i really need to learn a bit more about this and get my hands dirty in it
13:22:57JdGordonthis could be interesting.. autoaccepting git deltas from trusted email adresses..
13:23:02GodEaterJdGordon: wouldn't be much point given I don't have commit access to rockbox :)
13:23:43JdGordonscorche: he says git is better for everyone... so he shold be able to say the best situation for us
13:24:01scorcheof course....i just wish there were more details
13:24:05*GodEater is actually having difficulty tracking down the big L's email address
13:24:23GodEateris he still at transmeta ?
13:24:33scorcheGodEater: he likely has many for each project
13:24:55Galoisthe video had an email address in the cover slide
13:25:00GodEateractually - perhaps emailing him is the wrong thing to do
13:25:01safetydanDoes git work on Windows? Judging from references to "hard links" when cloning repos I'm guessing not.
13:25:14GodEaterperhaps we should direct an email at the git mailing list (assuming there is one)
13:25:22linuxstb_GodEater: No, OSDL IIRC
13:25:25GodEatersafetydan: no idea
13:25:33 Quit linuxstb (Read error: 110 (Connection timed out))
13:25:33scorchedoesnt sound to bad an idea if you dont mind it being public
13:25:36GodEaterlinuxstb: thought so
13:25:49GodEaterdoesn't bother me, does it bother anyone else ?
13:26:26LloreanAsking about how appropriate git is for our project you mean?
13:26:44Galoisgit sort of works in cygwin, but it's painful
13:26:45linuxstb_GodEater: More than you wanted to know....
13:27:18GodEaterlinuxstb_: far more :)
13:27:22 Quit driver8` ("I said the dingo ate your baby")
13:27:31GodEaterLlorean: that yes was at your question by the way
13:27:32LloreanGodEater: I'd say go for it, but be sure to mention that we do distribute binaries, the majority of our users need binaries, so there's very definitely one final repository despite distribution.
13:27:42LloreanGodEater: I figured it was, I was just slow typing a response.
13:28:28GodEaterLlorean: it was the main part of the question I was going to ask :)
13:28:42 Join AceNik [0] (n=AceNik@
13:28:44safetydanGodEater: yes lack of Windows support is a bit of a downer. There are alternatives though like Mercurial that do work on Windows.
13:28:44scorcheGodEater: this is likely much more than he will want to answer (he is a busy guy), but also ask about how much load is ont he maintainer(s) of the head
13:29:17GodEaterscorche: going with the mailing list approach - so hopefully we'll receive answers from *someone* if not Linus himself.
13:29:23JdGordonhow well does it do deltas on binaries? we could set up a branch for the zips instead of serving them from the www server
13:29:24 Quit billytwowilly (Read error: 104 (Connection reset by peer))
13:29:28GaloisI thought rockbox itself didn't even compile in windows unless you used cygwin
13:29:38GodEater <−− mailing list archive for anyone interested.
13:29:54*amiconn wonders why hardlinks shouldn't work on windows
13:30:01scorchehe mentioned mercurial in the talk as another distributed system, but doesnt much into it
13:30:04LloreanGalois: Yes, but many people use TortoiseSVN or something similar, so they don't do anything *except* the compiling in cygwin
13:30:13amiconnAt least on W2K and higher they're definitely possible
13:30:23 Join billytwowilly [0] (
13:30:34safetydanamiconn, true, ntfs supports them
13:30:35scorcheit would be worth finding out if git is able to be gotten for cygwin as well (without compiling)
13:31:04KanashimiTake care everyone.
13:31:05safetydanscorche: Mozilla project went with mercurial after much discussion partly because of the poor Win32 support of git
13:31:11*amiconn uses them (so-called 'junctions') to have only one testfile dir shared between all his sims
13:31:15*JdGordon waits for bagder to come in a say "we are not bloody converting again"...
13:31:43scorchewell, Zagor started this whole talk! =P
13:32:18Zagorheh, are you still talking about this? :)
13:32:26safetydanrepository checkout with git-svn takes a while on my poor laptop and that's just for the tip. I'd hate to hink how slow a full history conversion would be.
13:32:27*Zagor whistles innocently
13:32:32 Quit Kanashimi ("-")
13:33:10*GodEater will post his question at pastebin or equivalent before he sends it.
13:33:27scorchesafetydan: yeah...that would be one of the pitfalls of git-svn....honestly, provided the load isnt too much ont he maintainer(s) of the head, and we handle it right, i dont see a problem with going full git
13:34:06Zagorif we convert, we'll do it fully. no half-measures.
13:34:21Lloreanscorche: I think it'll slow down the 'commit' (pull?) rate quite a bit.
13:34:35AceNikwould anyone tell m , how to redesign a plun originally made for another player ?
13:34:43LloreanUnless there's a solid group of maintainers of the live branch around the clock
13:35:07LinusNAceNik: a what?
13:35:12JdGordonLlorean: I would guess anyone with current access would become a maintainer? I couldnt see how i would work without that
13:35:12LloreanLinusN: Plugin maybe?
13:35:18linuxstb_AceNik: You mean how to make a plugin which works on one Rockbox target work on another target?
13:35:27 Quit spiorf (Read error: 110 (Connection timed out))
13:35:47 Join spiorf [0] (
13:35:51LloreanJdGordon: That somewhat goes against the philosophy of git, though.
13:36:01 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
13:36:02dionoeaamiconn: i've never tried it on archos.
13:36:22JdGordonLlorean: no, i dont think so, the article I read says git has a way to push your commits up
13:36:26 Quit spiorf (Client Quit)
13:36:38 Join spiorf [0] (
13:36:49LloreanJdGordon: It has a way to push, as I said, it's against the philosophy, not necessarily the capability
13:36:50AceNikso sorry my apologies
13:37:23*JdGordon wonders how git blame is supposed to work if there is only 1 person doing real commits
13:37:46AceNikok more precise & correct : how o you resize & recode a plugin designed for another player , to work on the desired player
13:38:00LloreanAceNik: Do you know C?
13:38:17ZagorJdGordon: those "real commits" are tagged with the developer name, aren't they?
13:38:17AceNikyup a little
13:38:29Zagorisnwhat you
13:38:40LloreanAceNik: You basically need to open up the source for the plugin, and find the player-specific bits (screen positioning, keymaps) and write new ones for your player.
13:38:48JdGordonZagor: not by the maintainers name?
13:38:50 Quit Arathis ("Bye, bye")
13:38:53Zagorisn't what you call "real commits" rather more like a "collection" of other peoples' commits?
13:39:10 Join Arathis [0] (
13:39:17JdGordonby real commit, I mean the equiv of svn commit
13:39:20linuxstbAceNik: Or make the code more generic, so it works on any player (e.g. position items in terms of a percentage of the LCD's width, rather than an absolute number of pixels)
13:39:46ZagorJdGordon: well there is no equivalent as I understand it, since there is no One True Repository
13:40:07AceNikLlorean:i have opened up the .diff file made as a plugin by someone, i uess there are keymaps first , then the code of the game exists , so if i just add a section to have the H10 keys will it work
13:40:14scorcheZagor: but, we would have to have a "head" that we use for binaries and such
13:40:19Zagoron the other hand I could be confused, I haven't studied git much yet
13:40:27amiconndionoea: Hmm. I think all plugins should work on all targets whenever that makes sense and is technically possible
13:40:36Zagorscorche: indeed. we pick a maintainer we trust. just like linux.
13:40:40JdGordonso going by that, doesnt it sound like the repo would never actually be updated by anyone except the mainter? hence git blame would be pointless other than showing when a line was chagd
13:40:54LloreanAceNik: Is your screen bigger or equal in size to the screen of the player it's meant for?
13:40:56linuxstbAceNik: Don't edit the diff, edit the source code after you've applied the diff. How much you will need to change to make it work on the H10 will depend on how generic the plugin is.
13:41:02AceNikthats what i thought , that all pluins would work , but there is some work with the resizing
13:41:13ZagorJdGordon: I don't think "git pull" changes the author tag (or whatever it might be called)
13:41:16amiconn(Example for #1 not applicable: iriverify only makes sense on irivers, example for #2 not applicable: doom won't run on archos)
13:41:26LloreanZagor: I think we'd need a minimum of about 5-8ish maintainers.
13:41:31JdGordonZagor: ok
13:42:05LloreanZagor: Just to be sure things don't slow down when that one maintainer's life gets busy, and so that things can happen during 'night' in their timezone.
13:42:13LloreanRockbox seems to be characterized by lots of smallish commits.
13:42:20JdGordonwe'd need more maintainers than that... does any of the current commiters not have a job?
13:42:29AceNikfirstly there aint no source right now , it a plugin game called spider solitaire & the author made a .diff nopthing else , it might be making files i guess
13:42:30ZagorLlorean: yes that is a question mark about git. does it, in essence, make a bottleneck of a single maintainer
13:42:49AceNikcheck this out
13:42:50Zagorthat may be fine for linux, but I don't think we'd want that for rockbox
13:42:50scorchei dont see an issue with lots of maintainers on the head (and would probably prefer it), but i still need to figure out exacttly how much load we are talking about
13:43:08Zagorscorche: load?
13:43:11 Quit DerPapst (Read error: 104 (Connection reset by peer))
13:43:32dionoeaamiconn: I agree. (I didn't code the plugin in the first place)
13:43:51JdGordonaccording to git can automatically accept incoming email changes, so if we used that we would only ned to restict the emails to known addresses, and as long as the change actually merged ok we wouldnt need maintainers
13:43:54scorcheas in how much work that(those) maintainer(s) would have to undertake....what they actually have to do specifically
13:43:57dionoeaIf one of my commits yesterday broke anything i'd be glad to fix it though
13:44:02 Quit GodEater ("CGI:IRC")
13:44:15 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
13:44:22ZagorJdGordon: yeah, that sounds just about like our current system
13:44:29GodEater <−− email I plan to send. Comments / suggestions please.
13:44:58Zagorbut then we're back to a One True Repo again, which I thought was antithetical to the git philosophy?
13:45:23ZagorI should probably read more about it before discussing...
13:45:41scorcheZagor: we will need a head repo regardless for binaries, as i said...this is more about how the developer is affected
13:47:03scorchepeople who read the logs sure will have an eyeful today... =)
13:47:10GodEaterindeed :)
13:47:26scorcheand those who dont, will likely be lost
13:47:36GodEateranyone like / dislike my email ? Do I hit send with it as is ?
13:48:06LloreanGodEater: I think it's mostly good, based on my limited knowledge.
13:48:23GodEatersent then
13:48:36scorcheGodEater: add to 1) that it is also for support purposes
13:49:17GodEaterscorche: I'll bear it in mind in case anyone asks ?
13:49:34scorchecan i have a link to the thread when i appears, so i can watch it?
13:49:46JdGordonpondlife: (2 huors later :p ) all this git talk got me to fix that problem with 4 lins of code :D new patch up i a min
13:50:34peturbah... why did I ever post that git talk link in here :/
13:51:01GodEaterpetur: I did it first :)
13:51:06 Quit safetydan (Remote closed the connection)
13:51:09NIXON-Trenti cant get any player to work on my photo ipod that allows for fast forward and rewinding ! :(
13:51:10GodEaterI think =/
13:51:19peturGodEater: when?
13:51:25AceNikcan someone help me with this sokban patch it was petur who started it =)
13:52:00JdGordonGodEater: i dont reall tihnk the binaries need is the thing stopping git for us
13:52:03GodEaterpetur: 08:58
13:52:14AceNikspider solitaire?
13:52:14*scorche apologizes for pointing the finger at Zagor
13:52:23 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
13:52:46NIXON-Trentanyone know a movie player for rockbox that allows for fast forwarding and rewinding of a movie otherwise the players are just pointless to watch full length films...
13:52:49GodEaterpetur: ah - my mistake - but I'd have posted it anyway :)
13:52:50scorchepetur: his mention was before you, but his was largely ignored
13:53:23*scorche shuts up
13:53:23LloreanNIXON-Trent: Watching a film *requires* fast forward and rewind?
13:53:27peturI wish I never started it
13:53:30LloreanStartling, I've never had to in the theatre...
13:53:37*Llorean learns something new every day.
13:53:45 Part AceNik
13:53:48GodEaterLlorean: hahahaha
13:53:51NIXON-Trentgotta love smarty pants :)
13:54:02LloreanNIXON-Trent: Fast Forward and Rewind will exist some day, mpegplayer isn't "finished" in terms of features yet.
13:54:13NIXON-Trentwell im not watching it on a large screen where i've dedicated 2hrs of my time to watch the movie
13:54:25scorchepetur: well, if it leads to a better "life", you can be the hero =)
13:54:28LloreanYou're perfectly welcome to work o n'em.
13:54:34Lloreanon 'em.
13:54:37NIXON-Trentits more hey im traveling on a train or car and can watch part of the simposons eposide..
13:54:50NIXON-Trenti would if i knew the language u guys code in.
13:55:22NIXON-Trenti spent 2hrs trying to get podzillas player working but its a peice of crap and i am still un able to load or use it to play any file :/
13:55:26scorcheNIXON-Trent: well, the simple fact is that coming here and complaining will not do anything about it...either learn C and program it yourself, or wait patiently
13:55:36NIXON-Trentdid i complain
13:55:41NIXON-Trentno i asked for an alternative...
13:56:07NIXON-Trentand i was replyed to with a uneeded reply.
13:56:22NIXON-Trent 21:53:16 ] <Llorean> NIXON-Trent: Watching a film *requires* fast forward and rewind?
13:56:31LloreanNIXON-Trent: You could our mpegplayer "pointless"
13:56:43LloreanI provided my opinion that fast forward and rewind were not, in fact, "necessary"
13:56:48NIXON-Trentrefering to my situation.
13:57:13NIXON-Trentbut as we continue this converstation i guess theres no help here to be found :/
13:57:23scorcheif you consider it "pointless", then dont use it and get a device that is designed for watching videos
13:57:24LloreanIf someone implemented fast forward or rewind, we'd included it
13:57:40NIXON-Trentipodlinux's mv player can..
13:57:47scorchethen use that
13:57:48NIXON-Trentits just i've found it a pain to get working
13:57:55 Quit linuxstb (Read error: 113 (No route to host))
13:57:57LloreanNIXON-Trent: And have you noticed any OTHER differences in their video player?
13:58:18NIXON-Trentit works
13:58:23scorchethen use it!
13:58:28NIXON-Trentand it stops, plays, fastworward and rewinds
13:58:34scorchethen use it!
13:58:41LloreanAnd it uses a different file format, so their code is useless to us.
13:58:48NIXON-Trenti'm trying to but having problems getting it to work on my ipod..
13:58:58scorchewell, go to them about that then
13:58:59GodEaterfirst reply is in, it's not that helpful though =/
13:59:02NIXON-Trentso because i like rockbox better decided to ask if there was an alternative.
13:59:05 Join linuxstb__ [0] (
13:59:11LloreanGodEater: Is there a threaded view of i t?
13:59:18 Quit ljcrabs2 ("Leaving")
13:59:20NIXON-Trentsorche keep out of it :)
13:59:30LloreanNIXON-Trent: Again, that sounds like you're assuming that if there was a better video player, for some reason we'd intentionally not include it?
13:59:45NIXON-Trentu dont need to be llorean's instant repeat.
13:59:51scorchei am not...
14:00:01GodEaterLlorean: the link I posted above is all there is
14:00:19LloreanGodEater: MIssed that, thanks, found it.
14:00:28NIXON-Trenti dont know but if ipodlinux can make seeking functions im suprised rockbox hadnt :)
14:00:30NIXON-Trentthats all
14:00:37LloreanNIXON-Trent: They use a different video format
14:01:20LloreanSo you can't compare them on the same grounds
14:01:21scorcheNIXON-Trent: we are a project which consists of volunteers only...not all of us have the time to implement it
14:01:24LloreanSeeking works differently.
14:01:44LloreanOur video format is more compressed and offers a lot of benefits, but seeking code takes different work.
14:01:46JdGordonanyone know how much disk space havig the entire hitory locally is going to take?
14:01:47NIXON-Trentscorche i do know that :)
14:02:05*JdGordon doesnt like the idea of gb's of data instead of the 60mb of source code
14:02:15JdGordonwell, 30 and the .svn dupe
14:02:56scorcheJdGordon: well, he mentioned it compresses it and that each branch is a tiny little file, but other than that, i dont know
14:03:39JdGordonanyone with a fast enough computer wanna pay guienea pig and import all our svn?
14:03:55NIXON-Trentclassify fast.
14:04:14scorchei will tomorrow, likely...just to play...i have been putting off sleeping ever since this started though
14:04:46JdGordonIf i had linux on my brothers comp id do it there... little buggers comp shits on mine :p
14:06:48 Join AceNik [0] (n=AceNik@
14:07:19AceNikhey guys what does "skipping trailing CRs " mean
14:07:23JdGordoncoliux, its faster :D
14:07:32*scorche shrugs
14:08:38 Quit linuxstb_ (Read error: 113 (No route to host))
14:08:56JdGordonjhMikeS: you round?
14:10:26amiconndionoea: Do you think xobox might work in pure b&w? Or does it need greyscale at least?
14:10:42 Join barrywardell [0] (
14:10:47dionoeapure b&w should work
14:10:51 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
14:11:03dionoeaYou basically would have black dots moving on a white background
14:11:08dionoeablack borders
14:11:26dionoeaand draw the player's cursor using some inverse color or something like that
14:11:41scorcheand if it comes to it, you can do something like bubbles did with different designs per group
14:14:05 Quit printfXh4 ("Leaving")
14:16:11*scorche finally gets a chance to go to sleep and seizes it
14:16:57pondlifeJdGordon: Sorry, back now. Still good to try the new patch?
14:17:16 Quit linuxstb__ (Read error: 113 (No route to host))
14:17:51JdGordonplease :)
14:18:52pondlifeOK, building...
14:19:20*JdGordon says bugger it... im importing svn into git
14:19:30JdGordon^ is my rough timestamp for the start
14:19:38pondlifeJdGordon: Nope, still broke :(
14:19:47JdGordon33/13500 commits done
14:19:54JdGordonpondlife: how?
14:20:16JdGordonsame problem?
14:20:25pondlifeProblem #1: Scroll down a folder - the cursor bar can't get onto the bottom 2 lines.
14:20:51 Join kaaloo [0] (
14:21:13 Part kaaloo
14:21:14pondlifeAh, is this intentional? It's just that it maintains 3 lines lookahead at the top and bottom.
14:21:37JdGordonyeah, we try to keep the cursor off the top and bottom
14:21:44pondlifeIs that new behaviour?
14:22:15JdGordonthe remote lcd seems to have problems in the main menu though :( missed that testing
14:23:03JdGordon253 commits done, 2mb disk usage
14:23:23pondlifeHow does it decide how much screen to "reserve"? It's quite a lot....about a third..
14:23:42JdGordonthats exactly what it should be
14:24:01JdGordonmin of 2 lines, up to 1/4 of the screen
14:24:28pondlifeI get 3 lines when moving up, or 2 when moving down.
14:24:40pondlifeWould prefer a max of 2..
14:24:45pondlifeOr not at all.
14:24:52pondlifeBut that's just me.
14:25:10JdGordonits better with a bit of a gap, but yeah, would be better if both were 2 or 3
14:25:53pondlifeAh, I've found a bug which *is* new though.
14:26:55pondlifeNavigate into a dir with only 3 entries, then into a subdir of that with lots of entries. Move down lota (so you're near the end) and press LEFT to go back up.
14:27:17pondlifeThe cursor bar end up off the top of the screen !
14:27:49amiconnThat dir backtrachas been broken for months. It used to work properly in the past :/
14:27:52JdGordonyeah, fixed that
14:27:53pondlifeYou need to make it scroll when in the subdir.
14:28:20pondlifeThis seems to be a new bug though. Wasn't in the 7th March build at least.
14:28:22amiconnBacking out of a subdir should place the cursor on that very subdir
14:28:35pondlifeIndeed. it does, but doesn't ensure it's visible.
14:28:38amiconnFor quite a while it doesn't do so anymore
14:29:01JdGordonthats fixed
14:29:09pondlifeNew patch?
14:29:40 Quit lids (Remote closed the connection)
14:30:13 Quit ompaul (Connection timed out)
14:30:13*JdGordon wonders why there is no files outside of .git yet after 500 commits...
14:30:16 Quit intgr (Read error: 104 (Connection reset by peer))
14:30:19 Join intgr [0] (
14:30:29JdGordonunless the cvs->svn import did a crazy amount of empty commits?
14:30:54GodEaterhmm. I'm not sure either reply I've had from the git list is much help so far.
14:31:00pondlifeJdGordon: Pop it on
14:33:16GodEaterJdGordon: are you trying out git-svn ?
14:33:44 Quit AceNik ("bye guys have fun , enjoy !!!!!")
14:33:48JdGordonGodEater: no, im importing the current tree into git to see what sort of size we are looking at
14:34:04GodEaterit wasn't that huge when I did it
14:34:11JdGordonits pretty fast :)
14:34:42JdGordon10% done
14:34:49GodEateryeah, I thought so too
14:35:20JdGordonso shuold be done mby midnight-ish :p
14:35:59*GodEater has no clue what the time is where JdGordon is
14:36:06pondlifeJdGordon: Much better
14:36:32GodEateryou must be doing something other than a "git add ."
14:36:38GodEaterit took less than a minute for me :)
14:36:46JdGordonGodEater: git-svnimport
14:37:08*JdGordon appologises to haxx for the download wastage
14:37:28GodEatercan't try that here - no proper intermaweb access
14:37:30pondlifeJdGordon: LCD seems ok too, is that menu fixed?
14:37:42JdGordonnot quite
14:37:43GodEater30MB repo when done with git add though
14:37:54JdGordonwhich is expected...
14:37:54pondlifeNo probs, lunch time for me
14:41:00*Llorean is tempted to make use of the word "allow" a bannable offense in the forums.
14:41:24LloreanWell, not really, but reading posts that say "Why doesn't Rockbox allow me to do Blah" gets irritating after a while
14:41:42GodEaterJdGordon: can you use git-svnimport on an already checked out local repository ? Or does it have to come from the svn master ?
14:41:43 Join Rob222241 [0] (
14:41:44LloreanLike it's a matter of that feature being out there, waiting, but Rockbox is standing in the way saying "No, it's there but you can't use it"
14:42:04JdGordonGodEater: from the master i assume, it pulls the whole history also
14:42:32JdGordondoing the test because apparently git stores that everywhere...
14:42:45JdGordonso want t see how much disk space we woiuld all waste
14:42:57 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
14:44:05LloreanWell, for those of us using the vmware image, it's a non-issue, I think. :-P
14:44:14LloreanThe image already reserves more disk space than we're likely to need for it.
14:44:32JdGordonits a non issue if you need 500mb extra disk space?
14:44:38 Join XbooX [0] (n=xboox04@
14:44:54LloreanThe virtual disk file is like... 2gb already, by default
14:45:11LloreanIf I need to use 500mb more of that already-used 2gb, it's not a major concern.
14:45:27JdGordonwhy is it so big?
14:45:33LloreanCouldn't say.
14:45:34GodEaterask scorche :)
14:45:44JdGordonthats crazy :p
14:46:54GodEaterLlorean: found a better view of the git list :
14:47:13LloreanGodEater: I thought you mentioned a second response
14:47:30GodEaterthough for some reason it's still not showing the 2nd reply I had, and he definitely copied the list on it...
14:47:43***Saving seen data "./dancer.seen"
14:48:09GodEaterI'll pastebin that too :)
14:48:26amiconnThe 2GB virtual disk is actually rather tight
14:48:37Lloreanamiconn: Really?
14:48:44LloreanI hadn't checked, but I'd assumed there was some elbow room.
14:48:52 Join lee-qid [0] (
14:48:54amiconnTry building 10+ sims, enable ccache etc....
14:49:13 Quit GodEater ("CGI:IRC")
14:49:14amiconnI'd rather reserve 5..10GB for a full rockbox dev environment
14:49:26 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
14:49:28amiconnAs Zagor said, disk space is cheap...
14:49:33*GodEater would just rather use linux
14:50:26*Llorean still keeps delaying installing Linux on his laptop.
14:50:29amiconnThe OS doesn't matter here. The VMware image _is_ linux
14:50:43LloreanI wouldn't mind reserving 5 or even 10gb for it either.
14:52:05GodEaterbut you're having to "reserve" space with vmware
14:52:17GodEaterif the entire machine is running the OS, you don't have to reserve it
14:52:22GodEateryou can use the entire disk if you need to
14:52:25GodEateror not if you don't
14:52:41LloreanMy real problem is hardware I haven't tracked down a way to use under linux yet.
14:52:58GodEaterlaptops are a PITA like that =/
14:53:09Galoisyou could use linux on a second machine ...
14:53:11JdGordonLlorean: really new or really old hardwrae?
14:53:42LloreanJdGordon: HTC Faraday / 2125 used over USB for internet connection while travelling.
14:53:57LloreanSo, fairly new, but it's also windows mobile which makes things more exciting.
14:53:59*GodEater pretends to know what that is
14:54:08LloreanGodEater: 's a phone
14:54:14Galoismy verizon phone over usb works fine in linux
14:54:51LloreanI imagine mine might, I just haven't gotten around to putting together a liveCD of some sort and testing to be sure.
14:55:08LloreanBut I haven't found any specific accounts of users with it yet either.
14:55:20*petur made the switch to linux on his new laptop
14:55:29*GodEater needs to update his PS3 linux install to get the wifi working
14:55:42amiconnGodEater: The default for workstation vmware disk images is to grow. For some reason our dev image uses the reserve-all type...
14:56:08GodEateramiconn: isn't that because it was created in VMWare server ?
14:56:30amiconnEven the server allows for growing images, it's just not the default there
14:56:46JdGordon23% done....
14:56:50amiconnOnly ESX server doesn't, but those images aren't compatible with workstation or player anyway
14:56:53linuxstbamiconn: Were you working on building your own vmware image?
14:57:10amiconnI still have 2 problems with it
14:57:33amiconn(1) Even though it doesn't have X11, it's bigger than the one we're currently distributing
14:58:00JdGordonstart with a ultra-minimal slackware setup :)
14:58:01amiconn(2) For some reason building the html manual fails, although the same thing works on my full fledged debian vm
14:58:29amiconnBut it does build all targets and pdf manuals, and it does build win32 sims
14:58:42GodEaterESX Server does allow for image growing, it's just not compatible with Server / Workstation because it requires a SAN
14:59:05 Quit Rob2222 (Read error: 110 (Connection timed out))
14:59:09 Join AceNik [0] (n=AceNik@
14:59:17JdGordondoes anyone want to test the fix for the wierd sansa accel issue before i commit? (someone with an ipod would be excelant..)
15:00:13ZagorLlorean: phones just work like old dial-up modems. you say "ATDT*99#" to it and you get a ppp response back.
15:00:28Zagorso I'd be surprised if the HTC doesn't work
15:00:28GodEatergood ol' Hayes command set
15:00:29LloreanZagor: Ah, interesting.
15:00:38GodEaterthose were the days :)
15:00:45LloreanZagor: I had to jump through about 50 driver hoops to get it working in windows, it'd be lovely if it's simpler under linux.
15:00:50GaloisI bet it is
15:00:53Galoismy phone doesn't work in windows
15:01:09GaloisI mean, I think it might work if I jumped through 50 driver hoops, but fedora has the drivers built in
15:01:17ZagorLlorean: first test is to plug it into a linux box and see what it is identified as
15:01:38Galoisyou hook up the phone and /dev/ttyACM0 pops up, then you pretend it's /dev/modem like Zagor said
15:01:55Galoisthe only difference is it's faster than dialup
15:08:01JdGordoncan someone with an ipod plase check and tell me if it causes/fixes any scan accell issues?
15:11:50JdGordonpretty please....
15:16:56GaloisJdGordon: what am I supposed to be looking for after applying the patch?
15:17:12 Join darkless [0] (
15:17:25JdGordonanything odd in the lists when scrolling
15:17:33JdGordonwierd acceleratino
15:20:20GodEateramiconn: from what I can see, you can specify that disks grow in vmware server, but still only up to a given size. So you might as well pre-allocate them as it's faster for the VM when it's running.
15:20:43 Join lids [0] (i=lds@gateway/tor/x-d4a50084e64c3cbc)
15:21:26GodEater3rd reply on the git list. Perhaps the most helpful so far.
15:22:03GodEatersuggestion is to look here for how to set up a "central" git repository :
15:24:44 Part XbooX
15:25:43dionoeaGodEater: I doubt that disk growing has a noticeable impact on performance in vmware. (Since it doesn't grow very often)
15:26:07GaloisJdGordon: it seems to behave the same as before
15:26:16JdGordonok great
15:26:20JdGordonwill commit soon then
15:26:53*petur wonders if "same as before" means with the same bugs :p
15:27:32Galoisthat's kind of what I was asking, because I wasn't aware of any bugs with the old behavior
15:28:47linuxstbJdGordon: I can't test now (but will later), but does it still accelerate on the ipod when moving the wheel slowly?
15:28:57JdGordonthats what im asking about
15:29:04JdGordonit shuoldnt...
15:29:51linuxstbGalois: What happens when you scroll slowly - e.g. one line every second or two seconds.
15:30:03Galoiswith the patch, scrolling once around the wheel advances exactly 18 lines, whether it's fast scrolling or slow (ipod nano)
15:31:15Galoishere we go, let's try changing the scroll speed in settings
15:32:41Galoisoops, wrong meaning of scroll
15:34:16Galoisthis is not exactly scientific, but if I scroll super slowly, one line per two seconds like you suggest, it does scroll more lines
15:34:22Galois21 lines per revolution instead of 18
15:35:44JdGordondamn, ok
15:35:53 Join webguest08 [0] (i=3bb72d9a@gateway/web/cgi-irc/
15:36:18webguest08I have a problem
15:36:40peturjust ask the question
15:36:40 Quit webguest08 (Client Quit)
15:36:56peturalso fine...
15:37:05 Join saratoga [0] (i=9803c50e@gateway/web/cgi-irc/
15:37:17JdGordonGalois: are you leaving your fingure on the wheel when you do this?
15:37:36GaloisJdGordon: I just tried it without the patch, and it's *much* worse without the patch than with the patch
15:37:37JdGordoncan you tes releasing?
15:37:45JdGordonok great
15:38:07Galoiswithout the patch, it skips lines when I scroll slowly
15:38:18JdGordonlinuxstb: that sound alright to commit?
15:38:32Galoiswith the patch, it doesn't skip any lines, but it still changes lines at a slightly faster rate
15:38:52JdGordonas long as its no skipping when it shhuoldnt
15:39:03JdGordonwhat about going round quickly? does it skip as expected?
15:39:14Galoisgoing round quickly is the same either with or without the patch
15:39:23JdGordonk great
15:39:27Galoisthe skipping lines stuff only happens without the patch and if I don't release my finger
15:39:27*JdGordon commits
15:40:24linuxstbJdGordon: I'll let you know later when I test it ;)
15:41:25linuxstbJdGordon: But could you explain the algorithm in the commit message?
15:41:52JdGordonthe action code now checks the time difference
15:41:52JdGordonbetween events to decide if its a repeat or not
15:41:57JdGordongood enough?
15:42:45linuxstbSo it ignores the "repeat" flag?
15:43:02JdGordonit never checked it... it always went by the action value, not he button value
15:43:33JdGordonim not 100% sure this doesnt break other code though...
15:43:35*JdGordon checks
15:44:14JdGordonit wont :)
15:44:36linuxstbI'm not sure I understand - is the action code (i.e. the code to turn button events into actions) checking the time difference, or is the list code?
15:44:41JdGordonand from action.h... "#define ACTION_REPEAT 0x2 /* action was repeated (NOT button) */"
15:45:44JdGordonlinuxstb: both, the list wont unless the action was repeated, the list then checks to see if its accelerating yet, and if it is, if its time to increase the skip
15:47:07 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
15:47:34linuxstbOK, so back to my first question - is the time difference between button events the only way to trigger a repeat action, or is the repeat flag from the button driver also used?
15:47:37Bagderone wifi router dies, another one starts
15:48:19JdGordonlinuxstb: only the time diffrence (assumign the 2 actions where the same value)
15:48:41JdGordonpressing stop and left (fas enough) would cause a repeat
15:48:45 Quit AceNik ("bye guys have fun , enjoy !!!!!")
15:48:51JdGordonwould appear as a repeat
15:49:35linuxstbOK, thanks And my last question is what is the time difference that triggers an action repeat?
15:49:39 Nick idnar_ is now known as idnar (n=mithrand@unaffiliated/idnar)
15:51:00JdGordonthat might need to be tweaked...
15:51:07linuxstbI think that will work better on ipods (ignoring repeat events), and it's easy to accidentally untouch the wheel when scrolling.
15:51:22linuxstb^because it's easy
15:51:48JdGordonok, so commiting...
15:52:02 Join AceNik_ [0] (n=AceNik@
15:55:15 Part AceNik_
15:55:30 Join AceNik_ [0] (n=AceNik@
15:55:30 Quit saratoga ("CGI:IRC (EOF)")
15:55:32*amiconn wonders why we need an extra repeat mechanism
15:55:54amiconnWouldn't it be sufficient to change the way the ipod wheel fires repeat events?
15:58:54 Quit maffe (Remote closed the connection)
15:59:00 Quit ompaul ("reboot, rebuild, reinstall, reconfigure")
16:00:38JdGordonamiconn: without it in the action code, the acceleration in the lists isnt possible, unless we check it there also, so then there is no point moving it
16:01:04 Part AceNik_
16:01:13JdGordon... and it was the sansa wheel which brought this bug up.. not the ipod
16:01:47JdGordon52% completed... looks like im sleeping with the comp on tonight :'(
16:03:09*Nico_P had a long read in today's logs
16:03:14Nico_Pinteresting one too...
16:03:25 Join Rincewind [0] (
16:04:02Nico_PBagder: thoughts about git ?
16:04:13Bagdernot really
16:04:22 Quit Entasis (Read error: 104 (Connection reset by peer))
16:04:23Bagderother than it uses libcurl so of course it is cool ;-)
16:04:40JdGordonwrong question.... thoughts about possibly moving from svn->git
16:04:48Bagdersame answer
16:04:50amiconnJdGordon: How so?
16:04:58JdGordonhow so which?
16:05:00BagderI don't have thoughts
16:05:01Nico_PBagder: how about LinusT bashing SVN ?
16:05:06Bagderwe're not linux
16:05:11Bagderwe don't develop like they do
16:05:19Bagderlinus bashed it long before it even existed
16:05:23Bagderand continues to do so
16:05:28Bagderno surprise there to anyone
16:06:02Bagdernot saying git is bad, just saying that linus t's opinion on svn is nothing to care about imho
16:06:29amiconnJdGordon: The repeat stuff... I take it that the button repeat is enough for list acceleration on button-only targets?
16:07:02Nico_Psure... I just find funny how he likes to bash some projects
16:07:12Bagderhe likes to be a prick
16:07:23JdGordonamiconn: no, the list doesnt deal with buttons at all, only the ACTION_STD_NEXT, PREV, so either it checks the time diference there, or relies on something else to do it.. i.e the action code
16:07:37BagderI find his attitude and approach annoying in other projects than linux itself
16:08:22Nico_PI can understand that
16:08:25 Quit Zagor ("Client exiting")
16:08:36Nico_Pmaybe one day he'll bash rockbox :)
16:08:47BagderI'm sure he would if he ever used
16:08:56JdGordonhe can be added to the official rockbox hitlist :D
16:09:15Bagderjust look at his language and style when he attacked the gnome project
16:09:27Bagdertotally out of order
16:09:49Nico_Psomething along the lines of "interface nazis", right ?
16:09:55JdGordonwhat? giving them patches? (alright, he knew they wouldnt be accepted.. but stll)...
16:10:10Bagderusing lots of foul language, cursing, calling people names etc
16:10:22Bagderbut sure, he did provide a patch
16:10:54GodEaterhis approach may be off, but it doesn't make him wrong =/
16:11:24Bagderperhaps, but I would've just added him to my ignore list if someone came doing that on one of the projects I cared about
16:11:33GodEaterI'd probably have done the same
16:12:00GodEaterI share some of views of gnome - I don't like that so much of it's interface is locked down and very difficult to change.
16:12:12GodEaterconversely however, I don't like KDE - it goes too far the other way
16:12:30GodEaterbut I certainly wouldn't have taken the tack he did with talking to the devs about it
16:12:37GodEaterhe is a little arrogant
16:12:54JdGordonwas he baited into that argument? or was that just him being a dick?
16:13:18GodEatermainly him being a dick
16:13:36 Join arkascha [0] (
16:13:39RincewindGodeater: there is gconf-editor on gnome that can change almost anything, it is just a little bit hidden
16:13:57 Part arkascha ("Konversation terminated!")
16:13:59JdGordonpondlife: back from lunch yet?
16:14:00GodEaterRincewind: that's sort of my point. It's not very well documented.
16:14:11*Nico_P is a happy KDE user
16:14:29Rincewindtrue, I only found it a few weeks ago (using linux for half a year now)
16:14:31*JdGordon doesnt think anything linux/foss is documented properly
16:14:32GodEaterI'm aware of it of course - I just don't think you should have to use some odd esoteric config editor to change some apps method of data display
16:14:34Nico_Pbut I agree he was pointlessly hard on gnome
16:15:26BagderJdGordon: curl is! ;-)
16:15:40LloreanJdGordon: Rockbox is FOSS, and it's pretty well documented these days.
16:16:32JdGordonpresent company excepted then :)
16:17:12JdGordonand i disagree Llorean, but I cant write user manuals for shit (let alone cmment my code), so ill be quiet
16:17:24 Quit Rob222241 (Read error: 110 (Connection timed out))
16:17:37LloreanWell, I use "well documented" in a relative context, in comparison with both FOSS and commercial products.
16:18:02LloreanI'm pretty sure it's impossible to write a user manual that actually qualifies as "good" once the complexity of what you're describing gets past the intuitive point, and at the intuitive point you don't need one. ;)
16:18:53*JdGordon fairly sure at least one of his past lecturers will disagree.. but he isnt here
16:20:20*JdGordon itching to commit this list patch, FS #7258 needs testing :)
16:20:55pondlifeJdGordon: Yes
16:21:10pondlifeHang on, will test again...
16:21:53JdGordoncant actually remember if anything changed since you left though :p
16:21:55pondlifeHmm, that's the same patch isn't it.
16:22:19pondlifeLooks like it... was working ok for me with a basic test...
16:22:21Nico_PJdGordon: what do you want tested ? ie what should I look for ?
16:22:30JdGordonNico_P: any wierdness in the list
16:22:35JdGordonnice n vague :)
16:22:37Nico_Pok I'll test
16:23:17pondlifeYep, that third one is the one I tested
16:24:26Nico_PJdGordon: how's the SVN -> git going ?
16:24:26 Join netmasta10bt [0] (
16:24:31JdGordonso the only thing that doesnt work is the first item in multi-line lists is shown on the last line when you press up?
16:24:49JdGordonNico_P: almost 60%
16:24:51JdGordonafter 2 huors
16:25:00Nico_Phow much space ?
16:25:17Nico_Pwow that's small
16:25:28 Quit barrywardell ()
16:25:45JdGordonI have no idea how much data im pulling from the server which is a bit annoying, free time only just started :p
16:26:15Nico_PI'm starting to be really interested by git-svn for my GSoC work
16:26:32Nico_Psvn merging is scaring me a bit
16:26:59JdGordonyou still have the same headache of fixing conflicts...
16:27:19*Nico_P just got a super-weird fading pink screen on his gigabeat
16:27:39dionoeaPSOD ?
16:27:53Nico_Phaha not far from that
16:28:25Nico_PJdGordon: I can go below the last item in the main menu... I then get "undefined instruction"
16:28:37JdGordonhod you do that?
16:28:54Nico_Pnot hard... go to the last item and press down :)
16:29:07Nico_Por maybe it needs to be with a repeat action
16:29:59JdGordon! hadnt seen hat before
16:30:05Nico_Pnope, no need for that. I go to the last item. press down once: nothing is selected. press down once more: undefined instr
16:30:23Nico_Pgood thing I tested then :)
16:30:33JdGordononly on lists shorter than the lcd
16:31:25Nico_Pyes. the rest seems fine to me
16:32:09markunNico_P: sometimes when rockbox on the Gigabeat crashes the screen is no longer being updated and then you get a very weird fading screen
16:32:30 Join barrywardell [0] (
16:32:34dionoeawin 36
16:33:11Nico_Pmarkun: it's the very first time I've had that. and there's been a commit affecting the LCD code very recently... *maybe* it's related ? :)
16:35:35markunno, I don't think so
16:35:53markunI had it many times in the early days of the port
16:35:57 Join aliask [0] (
16:36:37Nico_Pit was a scary experience
16:36:40*Nico_P is still shaking
16:37:10 Quit barrywardell (Client Quit)
16:37:47 Join barrywardell [0] (
16:40:43pondlifeJdGordon: I think I just saw the same issue on the H300 sim... where you can scroll off the bottom, right?
16:41:19JdGordonyeah, apparnetly I never got it because the root menu exactly fits the screen with the test font
16:41:41pondlifeI'm usingthe Settings menu (it's a bit shorter)
16:41:53pondlifeWeirdly the LCD appears to wrap ok
16:41:59 Join Guile` [0] (
16:42:00pondlifei.e. the remote
16:42:34JdGordonall fixed
16:42:54pondlifeNew patch, or are you itching to commit anyway? ;)
16:43:00 Join mpeccorini [0] (
16:43:48JdGordonpatch online
16:44:16JdGordonnot commiting untill im sure everything i fixed, I and everyone is sick of commits to list.c :p
16:45:06pondlifeGood !
16:46:35 Join toffe82 [0] (
16:47:47***Saving seen data "./dancer.seen"
16:47:47pondlifeWell that seems ok, but then I never spotted the last problem!!
16:48:13pondlifeHow do I display a multi-line item?
16:48:28pondlifeDisk info?
16:48:30JdGordonid3 viewer
16:48:34JdGordonor partition info
16:50:13pondlifeLooks ok
16:51:02JdGordonit doesnt work on the remote wen there is only 3 lines, but that doesnt matter i tinhk
16:54:18JdGordonso, good to go?
16:54:20JdGordonNico_P: ?
16:54:31Nico_Pyes ?
16:55:23JdGordonfoun any other problems?
16:55:40Nico_Pno. haven't tested the corrected patch yet
16:56:11Nico_PI'm trying to get a VB .NET app I coded some time ago to run with mono
16:56:52JdGordonok, no prob
16:57:46*GodEater had no idea Nico_P was such a masochist
16:58:20Nico_Pwhich part do you find masochistic ? coding in VB .NET or trying t get the app to run on mono ? :)
16:58:39GodEaterthe first part is mildy masochistic. The second part is just off the chart :)
16:59:01JdGordon1 years worth of commits to go
16:59:07Nico_Pwe'llsee :)
16:59:16Nico_Phave you ever tried it ? yes. Trying to get it to run under mono? No.
17:00:14peturJdGordon: the number of commits has been growing much higher in recent months comapred to the first years
17:00:18Nico_Pit's an useful program and I don't have windows anymore so I got curious
17:01:12JdGordonpetur: yeah, sounds about right, 10k commits up to about this time last year... 5k since then
17:01:15 Join jgarvey [0] (
17:05:02 Quit barrywardell ()
17:05:14JdGordoncan anyone help me with a really simple script to print the time when a program isnt runnig anymore?
17:05:42GodEaterprogram-you-want-run && date ?
17:06:07Bagderprogram-you-want-run ; date even
17:06:09GodEateror, if you want the time even if the program dies, then "program-you-want-to-run;date"
17:06:16*linuxstb was too slow
17:06:35JdGordoni mean, print the time when a diffrent program finshes....
17:06:47GodEaternot following
17:06:48JdGordonone thats already bee started
17:06:54linuxstbdifferent-program-your-want-to-run ; date
17:07:06 Part LinusN
17:07:17GodEateryou mean the one you want to check on has already been started JdGordon ?
17:07:28linuxstbIf the program is modifying files, you can just check the timestamps of those files.
17:07:45GodEaterso you're going to need to check the output of ps, running periodically
17:07:50Bagderstrace -p [pid] ; date
17:07:52GodEaterfor the absence of said program
17:07:56Bagderbut it'll take cpu time... :-)
17:07:57GodEateryeah, that'd do
17:07:57JdGordoncould do that, but too hard, dunno what files its going to leave behind
17:08:09GodEaterbagder's is simplest
17:08:19linuxstbIs it too late to stop the program and start it again?
17:08:28GodEaterpossibly you should pass some more arguments to strace so it doesn't pick everthing
17:08:45GodEaterand I''d redirect the output of the strace to /dev/null too
17:08:50Nico_Plinuxstb: I suspect the program is the SVN to git conversion so probably
17:09:01Nico_Pam I right ?
17:09:04JdGordonNico_P got it in one!
17:09:05GodEatergood guess I think Nico_P
17:10:36linuxstbBut yes, the suggestion of "strace -p [pid] &> /dev/null ; date" will work...
17:11:00Galoisyou could probably get away with typing date in the terminal from which you launched the git-svnimport
17:11:11JdGordonwill that eat lots of cpu?
17:11:16Galoisbecause, you know, terminals are smart and have buffers
17:11:32JdGordongood idea :)
17:12:10GodEaterhow about "strace -enone -p [pid] >& /dev/null ;date "
17:12:55GodEaterGalois: if he has a decent bash prompt, he wouldn't even need to type date :)
17:13:28*GodEater has a kick ass bash prompt
17:13:33Galoisso you're suggesting ^Z; export PS1=something decent; fg
17:13:48GodEaterer, not quite no :)
17:13:52JdGordonGodEater: yeah, I thought about fixing that, but does it take affect without reloading?
17:14:13GodEaterwell you could do as Galois says - but the ^Z would stop the process running while you make the change
17:14:25GodEaterI'm not sure how important the total running time is to you
17:14:36JdGordonnot that important...
17:14:42JdGordonwhats 30s when its gone 3 hohuors alreadt
17:14:58GodEaterI could email you my bash prompt script ?
17:15:06JdGordonanyway,. got the strce going, and the buffered date... (some hoping it doesnt ask for input :D)
17:15:16JdGordonna im right
17:15:48GodEaterstrace doesn't ask for input, it's output only
17:15:48*JdGordon going to sleep
17:15:58JdGordonno, the git import program
17:16:14JdGordonpondlife: if you spot any other bugs, lemme know in the tracker, otherwise ill commit in the mrning
17:16:20Galoisluckily, git-svnimport doesn't
17:16:36 Nick JdGordon is now known as JdG|Zzzz (n=jonno@rockbox/developer/JdGordon)
17:17:17*GodEater 's fancy bash prompt is available to anyone who wants it - it looks best with a LC_ALL set to your_lang.utf8
17:17:53 Join dster [0] (
17:18:22GodEaterand in a unicode terminal of course...
17:19:19dstercharger and usb pluged into my photo won't boot anything
17:19:29*Rincewind likes his prompt to be simple and small
17:19:44dsterany other ideas to what i can try?
17:21:05Lloreandster: Leave it plugged in two hours, turn on hold, turn off hold, then hold down menu+select for 30 seconds. Try the menu+select a few times to make sure you're not letting your fingers slip slightly sideways
17:21:52dsterWell, should I try to let it discharge before I try that?
17:21:54*GodEater is stunned
17:22:04GodEatera response from Mr. Torvalds himself!
17:22:05dsterAnd, you mean plug it in the wall with hold off at first?
17:22:19linuxstbGodEater: Was he reading our IRC logs?
17:22:45GodEaterno - it's the git email list
17:23:04GodEaterI posted a couple earlier
17:23:29*linuxstb scrolls up
17:24:18GodEaterit's a nice response too :)
17:24:30Nico_Pdamn the adresse is unreachable to m
17:25:14GodEaterhe didn't flame us :)
17:25:22Nico_PGodEater: would you mind posting the response on pastebin ?
17:25:34GodEaterNico_P: there's another link further up not on gmane
17:25:59GodEaterbut sure :
17:26:26LloreanGodEater: His response was actually pretty much advocating how I imagined we'd best make use of git anyway. Heh.
17:27:54RincewindGodEater: can you post the link to the original message, too?
17:28:08Nico_Pman that's a long response
17:28:27peturindeed - he took his time...
17:28:43GodEaterLlorean: indeed.
17:29:05GodEaterI'm responding so he doesn't think we're idiots. (i.e. on the checking in binaries question)
17:29:49 Join otih [0] (
17:31:55markunwhat kind of binary files do we have? Just the bitmaps?
17:32:30LloreanWell, the original post referred to the actual build binaries.
17:32:35Nico_PI think he thought we wanted to keep track of the rockbox bins
17:32:48LloreanBut it wasn't talking about including them, just the fact that we need to build ones frequently from a central repository
17:33:54GodEaterRincewind: original message is also pastebin'd much earlier in today's log
17:34:23Rincewindnvm, i searched around gmane and found it, thanks
17:35:57markunLlorean: yes, it looks like all of them interpreted the point about binaries wrong
17:37:37*GodEater hits send again
17:38:13 Quit atsea- (Remote closed the connection)
17:39:28*GodEater curses gmail once AGAIN for not doing reply-to-all by default.
17:40:19Galoisscrew that, it should have reply-to-list
17:40:31GodEatermy reply (for the impatient)
17:40:53GodEaterGalois: yeah or that. I've lost count of the times I've replied to a list message, only to find I only replied to the poster, not the list.
17:41:05GodEaterit makes me look like a complete moron (more so)
17:41:29GodEaternot that that will surprise Mr. Torvalds at all I suspect :)
17:41:53Galoismaybe someone should go to google and give a talk on email
17:43:58Nico_PGodEater: "I know you think everyone else is a moron" :D
17:44:17markunGodEater: maybe you could have done some top-posting to piss him off :)
17:44:22Galoisit makes sense if you watched the video
17:44:28Nico_PI did
17:45:24 Quit linuxstb (Read error: 113 (No route to host))
17:45:34 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
17:48:17GodEatermarkun: was very tempted ;)
17:48:18 Quit netmasta10bt (Read error: 104 (Connection reset by peer))
17:48:37*GodEater is off to the gym. Later all
17:48:42 Quit GodEater ("CGI:IRC")
17:48:54 Join atsea- [0] (i=ariel@gateway/tor/x-e0663a7d450b7f76)
17:49:04 Join Nosgoth [0] (
17:49:51Nosgothwhere can i ask for the wiki write permission?
17:52:01peturwiki name?
17:53:16peturNosgoth: what is your wiki name?
17:57:44 Join david1 [0] (n=david52@
17:59:14peturMartinTykal has write access
17:59:24Nosgoththank you=)
17:59:25 Part Nosgoth
18:03:48peturaustriancoder: the author of the wasabi stack is responding again :)
18:12:08 Quit Seed ("cu, Andre")
18:12:38 Join Seed [0] (
18:20:53 Part Rick|
18:27:15 Join saratoga [0] (i=9803994d@gateway/web/cgi-irc/
18:27:34 Join nls [0] (
18:27:57saratogalinuxstb: you around?
18:28:11linuxstbI am.
18:28:57 Join Psilonaut [0] (
18:29:26saratogai am really confused with a problem
18:29:53saratogaI have tracked it down to the MDCT stage, where it occasionally spits out a pair of completely wrong values, generally evenly spaced about the center of the block
18:30:18saratogahowever, I'm pretty sure nothing is underflow or overflowing
18:30:31saratogais there some other hazard to look for?
18:31:03saratogagiven that the fp version works correctly and the code is largely identical except for changing * to fixmul32
18:32:55 Quit lids (Remote closed the connection)
18:33:02linuxstbI'm not the person to ask about the gritty details of codecs. Do you get the same problem in both the sim and on a target?
18:34:01saratogasorry, this is just running from the command line
18:34:13 Join My_Sic [0] (
18:34:16saratogaon my linux machine, using a test driver
18:35:28 Join bluey- [0] (
18:35:56petursaratoga: you're supposed to work all summer on it, or did you plan a big vacation afterwards :p
18:36:15 Quit Rondom (Read error: 113 (No route to host))
18:40:01 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:41:48 Join lids [0] (i=lds@gateway/tor/x-3212f6cd90ce5785)
18:42:31 Join Xerion [0] (
18:42:45 Quit obo ("KVIrc 3.2.6 Anomalies")
18:45:19 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
18:45:52linuxstbsaratoga: Are you able to easily switch between using the floating-point MDCT and the fixed-point version?
18:47:49***Saving seen data "./dancer.seen"
18:52:08 Join Xerion_ [0] (
18:54:51nlsLear: did you post your gcc hack anywhere?
18:56:03nlsLear: do you think it would be a good idea to use it for building rockbox or was it just an experiment?
18:57:33LearWell, it only makes gcc generate slightly better (and smaller) code for us, but it was more than an experiment, I'd say. Should probably file it as a bug for gcc 4.3 though...
18:59:21nlsLear: That might be a good idea, could you post it somewhere so I can take a look, I've never hacked at gcc but is a little curious :-)
19:00:05 Join _Xerion [0] (
19:00:12 Quit Xerion (Read error: 104 (Connection reset by peer))
19:00:23 Nick _Xerion is now known as Xerion (
19:01:51LearI could send it to the developer mailing list, since the same "problem" might apply to ARM as well.
19:02:13nlsworks for me :-)
19:04:01 Join pepie34 [0] (
19:07:08saratogalinuxstb: yes I've been going through and comparing them sample per sample to find bugs
19:08:08 Part Llorean
19:08:35 Join Caskey-10 [0] (
19:09:47 Join Caskey-91 [0] (
19:10:22 Quit Siltaar (Remote closed the connection)
19:10:35Caskey-91yea quick question about Rockboy on the 5g ipod
19:10:47markun(oh no :)
19:11:28markunCaskey-91: well, let's get it over with..
19:12:20*linuxstb thinks Caskey-91 was lying about the question being quick...
19:13:14dionoeamaybe it was so quick that we didn't see it fly by
19:14:30linuxstbWhich reminds of Douglas Adams' quote regarding deadlines - "I love deadlines, especially the whooshing sound they make as they fly by"
19:14:37 Quit Xerion (" ")
19:14:49nlslinuxstb: just like the 3.0 deadline :-D
19:15:32linuxstbPrecisely that sound ;)
19:17:30 Quit Xerion_ (Read error: 110 (Connection timed out))
19:17:39peturlinuxstb: you'll not believe it, but I have that quote hanging above our planning chart here at work ;)
19:18:01peturmy boss didn't think it was funny
19:18:03linuxstbpetur: I thought it best not to mention it to my boss...
19:18:08 Quit davina_ ("xchat on Ubuntu 7.04")
19:18:32*petur is doing extra time at work :/
19:21:38 Part david1
19:23:37 Join bluebrother [0] (i=uLJv2J5q@rockbox/staff/bluebrother)
19:24:32 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
19:25:16 Quit linuxstb (Nick collision from services.)
19:25:18 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
19:25:31 Quit aliask ("ChatZilla [Firefox]")
19:27:15 Quit Nibbier (Read error: 113 (No route to host))
19:28:09 Join The-Compiler [0] (
19:29:29 Quit Caskey-10 (Read error: 110 (Connection timed out))
19:29:33 Quit Caskey-91 (Read error: 110 (Connection timed out))
19:30:50 Join Siltaar [0] (
19:35:20 Join lids_ [0] (i=lds@gateway/tor/x-456c3db35bad1876)
19:35:35 Join desowin [0] (
19:35:40 Quit lids (Remote closed the connection)
19:38:41 Join obo [0] (
19:38:52 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
19:39:00 Quit desowin (Client Quit)
19:45:47 Join Gwall [0] (n=Gwall@
19:46:55 Quit linuxstb (Read error: 113 (No route to host))
19:47:23 Join Buschel [0] (
19:51:54 Join Nibbier [0] (
19:58:15 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
19:58:38 Join linuxstb__ [0] (
20:05:19 Join kaaloo [0] (
20:05:26 Part kaaloo
20:07:16 Quit pepie34 ("Ex-Chat")
20:12:39 Quit linuxstb_ (Read error: 110 (Connection timed out))
20:17:04 Quit Psilonaut (Read error: 104 (Connection reset by peer))
20:18:14 Join bundik [0] (
20:20:20 Part bundik ("Leaving")
20:32:06 Join Rondom [0] (
20:32:45 Join kubiix [0] (
20:34:01 Join GodEater [0] (
20:34:38GodEaterwoohoo - internet at home again at long long last
20:35:13 Quit petur ("work-> home")
20:35:36LloreanGodEater: Linus T's second response is actually quite good overall too.
20:35:39 Part Gwall
20:36:01GodEateryes, I read it ob the bus on my way home :)
20:36:06 Join DerPapst [0] (
20:36:20*DerPapst is impressed
20:36:39GodEaterI'm being thrown off the laptop though
20:36:42DerPapst2 replies by Linus T o.O
20:36:42GodEaterbe back in mo
20:36:45 Quit GodEater (Client Quit)
20:36:49 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
20:37:07DerPapstand they were quiet nice too :D
20:39:14 Join ptw419 [0] (
20:43:04 Join GodEater [0] (
20:43:39GodEaterright back again
20:47:50***Saving seen data "./dancer.seen"
20:48:26GodEaterDerPapst: yeah, he didn't call me a moron once :)
20:49:34 Quit orbit7 (Remote closed the connection)
20:54:01 Quit amiconn (Read error: 110 (Connection timed out))
20:54:02 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
20:54:35 Quit Buschel (Read error: 104 (Connection reset by peer))
20:54:49 Quit Mouser_X (Read error: 110 (Connection timed out))
20:58:36GodEatermy poor harddrive
20:59:05 Join maffe [0] (
20:59:06GodEaterover a month since it last sync'd with portage - it's getting hammered
20:59:27 Join barrywardell [0] (
21:00:35 Join orbit7 [0] (
21:01:49 Join Buschel [0] (n=AndreeBu@
21:01:51 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
21:03:06 Join mareviq [0] (
21:03:46DerPapstGodEater: btw you called yourself a moron.. so he didn't had to do it anymore :P
21:03:56Nico_PGodEater: I managed to run VB .NET apps in linux :)
21:04:11Nico_Pnot all of them and not the main one I wanted, but still
21:04:48Nico_Pthe mono VB compiler seems very picky though, so I didn't manage to compile any programs
21:05:47GodEaterv. impressive Nico_P :)
21:06:30GodEaterDerPapst: I didn't think I called myself that - I think I just commented that he though everyone else was a moron. Not *quite* the same - but I see what you mean :)
21:06:36Nico_PI'm impressed by mono... very nice project
21:06:57Nico_Pand congrats on not being called a moron by LinusT :)
21:10:27Nico_PI've been plating with git-svn and I'm still finding it a bit hard
21:10:54LloreanWell, as was said, if we move to git, it's going to be all or nothing from the sounds of it.
21:12:29Nico_Pwell actually from what I understood, there's nothing preventing some of us from having git repos mirroring the SVN repo for a while, and pushing/pulling from each other
21:12:55Nico_Pit's far from optimal of course but it could work
21:13:34GodEateryep, there's nothing anyone can do to stop people with commit access from using svn-git
21:13:58LloreanIt might be a good way to both learn git some more, and get used to using it for better collaboration.
21:14:19LloreanIt might even be good for the SoC stuff, making it quite easy for the mentor to pay attention to the work being done by the student. ;)
21:14:32Nico_Pyes, there could be some separate git repos for experimenting
21:14:58GodEateras I said, I think dan_a already uses git-svn
21:15:00Nico_PLlorean: I was thinking of a SVN branch for that, but git does sound appealing :)
21:15:18LloreanNico_P: git would certainly make it easier, I think
21:15:46Nico_Pbtw you might be interested in knowing that I've been thinking about my project but haven't had time to actually start on it
21:16:08LloreanI was going to ask where you're standing right now. ;)
21:16:21Nico_Pmy shcoolyear isn't finished yet and I have a few exams left
21:17:20Nico_Pand I'm kinda maturing the ideas in my head, probably a bit like amiconn does with viewports :)
21:17:55Nico_Phopefully there'll just be less time between the idea and the code
21:18:33LloreanSounds good
21:19:22Nico_PI've been thinking about pondlife's BufferingApiProposal and discussing it with him a bit
21:26:48LloreanI've read a few logs of that in channel at least
21:27:27 Quit Nibbier (Read error: 110 (Connection timed out))
21:36:07 Quit The-Compiler (Remote closed the connection)
21:36:48 Nick linuxstb__ is now known as linuxstb (
21:41:09 Join perpleXa [0] (i=perpleXa@unaffiliated/perplexa)
21:41:37 Join Xerion [0] (
21:43:08Learlinuxstb: tried to download the demac tarball, but it ends up corrupted here. Don't know yet if it is a simple cr/lf issue...
21:44:59 Join Faemir [0] (
21:46:00 Quit tchan (Read error: 110 (Connection timed out))
21:46:38Learlinuxstb: He, had to run 'gzip -d' first, then 'tar xzf' worked...
21:46:49linuxstbLear: Yes, I've just noticed that myself...
21:47:40linuxstbI've had problems in the past with gzipped attachments on flyspray being gzipped twice. The file I uploaded was definitely only gzipped once...
21:47:56 Join davina [0] (
21:48:04linuxstbAre you using Firefox?
21:48:52linuxstbIf I download the file with wget, it's fine...
21:49:04linuxstbWith firefox, I end up with the double gzipping effect...
21:49:39 Join kaaloo [0] (
21:50:01LearFirefox, yes.
21:51:00Nico_Pis it on purpose that emacs is ugly ?
21:51:27Nico_PI just built v22.1 with hopes that the GTK interface would like nice, but it doesn't at all
21:53:13Nico_POK, looks like GTK has to be activated in the configure script. I'm shutting up and rebuilding
21:53:24 Part kaaloo
22:00:46 Quit bluey- ("Leaving")
22:02:25 Join BRi7Xxon [0] (
22:02:39 Quit alienbiker99 (Read error: 104 (Connection reset by peer))
22:02:54 Quit BRi7X (Read error: 104 (Connection reset by peer))
22:04:02 Join alienbiker99 [0] (
22:07:44 Join hostf4cekilla [0] (
22:07:52nlsLear: just rebuilt gcc with your "hack", everything seems to work nicely, some quick tests with test_codec shows no speed difference at least for vorbis.
22:08:24LearNo surprise there (about the speed difference).
22:08:57nlsquite a few of the codecs also dropped a few hundred butes in size
22:09:05amiconnLear: I think that if we want to use that patch, we should just increase the "rockbox patch number"
22:09:19amiconnI.e. this would become rockbox patch #2
22:10:08 Join TrueJournals [0] (
22:10:20amiconnDoes this also work for 4.0.x?
22:10:29LearMakes sense. But it would be a little confusing with #1 being for 4.0, and #2 for 3.4...
22:10:33*amiconn would like to switch to 4.0.x for coldfire as well
22:10:44*nls would like to handle the patch for building m68k-gcc on amd64 too
22:10:45LearHaven't looked, but I'd guess it would be similar, if not identical.
22:11:02amiconnMost things run at the same speed, and wavpack decoding becomes faster
22:11:27amiconnDidn't try any encoder though...
22:11:58LearNo, same patch wouldn't work, but manual modification is simple.
22:13:44 Part mareviq ("nie wszystko dioda co sie swieci...")
22:15:14bluebrotheris there any reason that keeps us from switching to gcc4 for coldfire?
22:16:31LearSlower code, I guess...
22:17:27 Quit ompaul ("reboot, reconfigure, retest, hope")
22:18:45 Quit _Veseliq_ (Remote closed the connection)
22:19:45 Join limbus [0] (
22:20:00limbushi all
22:20:53scorcheGodEater: the mail was a success =)
22:21:08nlswhy are we only building for the archoses with -Os and just -O for others? shouldn't -Os be both faster and smaller than -O?
22:24:04 Join jaczehack [0] (
22:24:53 Quit GodEater (Read error: 113 (No route to host))
22:27:01 Join GodEater [0] (
22:30:21 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
22:31:56 Join din [0] (
22:32:34 Quit rift ("d'accord")
22:34:47 Quit Thundercloud (Read error: 104 (Connection reset by peer))
22:35:14amiconnnls: Try -O2 or -Os ...
22:35:15 Join petur [0] (n=petur@rockbox/developer/petur)
22:35:25*amiconn expects crashes and/or failing builds
22:35:56 Quit jaczehack (Remote closed the connection)
22:36:06 Join Thundercloud [0] (n=thunderc@
22:36:23nlsamiconn: ah, ok, I have my last exam tomorrow, after that I will play around with it, thanks for answering my questions :-)
22:37:02amiconnIt was quite a bit of work to get the archos builds actually work with -Os (and it is the reason for rockbox patch #1, fixing a gcc bug)
22:39:34toffe82sorry for the silly question (I am not use to linux) but where I modify the path in linux, I got this message : The compiler you must use (arm-elf-gcc) is not in your path!
22:40:40 Join barrywardell_ [0] (
22:42:03limbustoffe82, IIRC, you got to edit (as root) the file /etc/profile
22:42:04mpeccorinitoffe82: you can change it in your .profile file /home/youruser/.profile
22:42:13limbusyou'll find PATH=...
22:43:21mpeccorinilimbus: I don't use linux but unix, doesn't linux have an independent .profile for every user?
22:43:40limbusI think one overloads the other
22:43:53limbusI'm just looking, but not finding either
22:44:00amiconnYOu can extend the path just for your user in the user specific profile
22:44:01mpeccoriniuse ls -la
22:44:14mpeccorinithe dot at the begining makes it "invisible" :p
22:44:24amiconnIt depends on the shell, for bash (which is standard on most distros) it's .bashrc
22:44:45amiconnJust put a line in there reading like:
22:44:54toffe82I am on ubuntu
22:45:04amiconnexport PATH=/path/to/your/arm-elf-gcc:$PATH
22:45:13toffe82sorry ;)
22:45:39mpeccorinitoffe82, try "ls -a" or files starting with "." will not be shown
22:46:08amiconnThen log off and log on again to activate the change, or execute the same command directly on the commandline
22:46:38mpeccorinior run ". ./.profile"
22:46:38 Quit Xerion (Read error: 104 (Connection reset by peer))
22:47:07 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
22:47:09mpeccorini". ./.bashrc" if you're running bash shell (per amiconn)
22:47:16peturin nautilus, press CTRL-H to view hidden files
22:47:54***Saving seen data "./dancer.seen"
22:48:27 Quit barrywardell (Read error: 113 (No route to host))
22:48:32 Quit petur ("switch!")
22:48:49 Join petur [0] (n=petur@rockbox/developer/petur)
22:50:25 Quit saratoga ("CGI:IRC (Ping timeout)")
22:52:29Learamiconn: patch is now on CrossCompiler wiki.
22:52:59limbusI want to add my server (at home, SDSL connection) to the rockbox build farm. just finished, and I think I am ready to receive the public key and
22:53:36limbusmayber Bagder can guide me through the rest of the process ?
22:53:50 Join Rob2222 [0] (
22:53:57limbushe's listed on the BuildServer wiki page
22:54:19 Nick Caliban_ is now known as Caliban (
22:54:41Bagderis it a fast machine?
22:54:52limbusIntel Core 2 Duo 1,86
22:55:06limbus1 GB ram
22:55:15Bagder is the script that should be named and get +rx permissions
22:57:48*petur kicks austriancoder awake
22:58:02Bagderlimbus: check the PATH at the top of so that they match your system. you have sdl installed too? then tell me your host name and I can do a first simple verification
22:58:22limbusI'll tell you as soon as the public key is in place
22:58:23austriancoderpetur: ping
22:58:28 Part TrueJournals
22:58:43limbusdoes check for sdl ?
22:58:55Bagdertry building a sim
22:59:01peturaustriancoder: petur/">
22:59:31austriancoderpetur: thanks
22:59:57peturaustriancoder: warning, the code builds but is a mess
23:00:21peturand the author of that code just returned my mail today
23:01:32 Quit Lear ("ChatZilla [Firefox]")
23:01:35austriancoderpetur: will start looking into a nice layer to access different usb chips...
23:03:06 Quit pondlife ("disconnected has pondlife")
23:03:54barrywardell_just's the work on the usb driver going?
23:04:07peturevaluating stacks
23:04:48peturand as a sideproject, I'm trying to get the usbotg driver running on h300
23:04:54austriancoderpetur: we will need more different layers i think. one for device part, one for host/otg part.. what do you think?
23:05:24peturforget otg, just host and device
23:06:07peturI need to look more at that code to see how it switches
23:06:09barrywardell_I looked at the lpc214x usb stack in the past
23:06:45peturbarrywardell_: there's a wiki page about the stacks austriancoder evaluated
23:06:58*barrywardell_ checks it out
23:09:47 Join pepie34 [0] (
23:10:19austriancoderpetur: md5sum okay? 8708f314c6849842190bd69a6058f757
23:10:27austriancoderI have problems to open the zip
23:11:26peturI'll email it, the webspace is probably full
23:11:47 Quit davina (Remote closed the connection)
23:11:51 Join entheh [0] (
23:11:54 Part perpleXa
23:13:32peturaustriancoder: check your gmail box
23:14:18austriancoderpetur: thanks.. now it works
23:15:03*petur has too many devcon pics on his webspace... damn ISP giving so little
23:17:20 Quit Faemir ("I could put something witty here, but no...")
23:20:09 Quit Thundercloud (Remote closed the connection)
23:22:38 Quit Rondom ("Ex-Chat")
23:24:22 Quit pepie34 ("Ex-Chat")
23:26:47 Part nls
23:27:21 Join miepchen^schlaf [0] (
23:27:23 Quit Siltaar (Remote closed the connection)
23:30:36Bagderso, anyone feels like doing a commit? ;-)
23:30:50peturgive me 10 minutes ;)
23:31:07Bagderyou can have 8, no more
23:31:26limbusthen he'll do in 6 :p
23:31:59*petur slaps Bagder with a large something
23:32:18*joshin knows that 10 minutes to a programmer is twice as long as a minute is to a wife/girlfriend getting ready to go out
23:32:47BagderI think it's more like a factor pi
23:35:36 Quit jgarvey ("Leaving")
23:35:50 Join Soap [0] (n=Soap@rockbox/staff/soap)
23:38:53peturin 8 :)
23:41:17peturwow... are we down to 4 minutes for all builds?
23:42:58Bagderindeed impressibe
23:43:03Bagderuh impressive
23:45:04peturlooks like it is hanging somewhere?
23:45:18Bagderyeah, godeater...
23:45:29Bagder"since 348"
23:46:43 Join Cardini [0] (i=189fceb0@gateway/web/cgi-irc/
23:47:14Bagderone of these days I need to write a fix for that
23:48:07limbusrestart ?
23:48:24BagderI don't have any good way to just nuke one single server's works
23:48:24CardiniI have a problem: my player is stuck on the RockBox start-up screen, with the dialogue "Committing database [1/8]." This seems to be because I chose the "Initialize Now" option in General Settings>Database. Model is Sansa e270 6GB, version is current.
23:48:39peturheh, you're going to rewrite godeater? Good luck :p
23:49:00Bagdercan't be that hard, can it? ;-P
23:49:27limbusBagder, cpu-load still hilarious (less 10%)
23:50:34Nico_Plimbus: are you doing "make -j n" ?
23:50:34LloreanCardini: How long has it been there, and how many files do you have?
23:51:09limbusNico_P, -j 2 on a Intel Core 2 Duo
23:51:26limbusnot sure whether I can set 4
23:51:29Nico_Pant you only have 10% CPU load ?
23:51:41Bagdergrrr, after 190 seconds(!) one build is left... on godeater
23:51:52limbuseat 'em !
23:51:57Nico_Pcan't you just diable him ?
23:52:14Nico_Prrrr... I need to learn to type
23:52:19BagderI need to, yes
23:52:19 Quit orbit7 (Read error: 104 (Connection reset by peer))
23:52:23CardiniLlorean: It was up for 15 minutes before I reset the player, and additional 21 minutes after restarting. I then reset it again and am charging it using the original firmware. Files? About 550 music files, and I think the total is about 920.
23:52:40BagderI wish I knew why it happens...
23:53:31limbusNico_P, precisely "-j2" whitout spaces
23:53:50limbusccache enabled
23:54:26*amiconn whispers "speculative parallel builds"
23:54:43limbuson one of the cpus
23:54:50 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
23:54:52Bagderamiconn: that's not really the answer to this problem though
23:55:11Bagderin fact
23:55:15amiconnNo, but it reduces the severity problem
23:55:16peturauto-kill after xx time?
23:55:21Bagderit makes this problem harder
23:55:25amiconnInsert "of the" somewhere..
23:55:41Bagderyeah, I need to kill jobs taking more than N secs
23:56:05Bagderbecause they're dead
23:56:34Bagderand if we parallel build, another server will build and provide a zip
23:56:43Bagderor might at least
23:57:13amiconnThe idea of speculative parallel builds is that in the end, several servers are building in parallel. When the first server returns the finished build, all parallel versions of that build would be killed
23:57:24Bagderyes of course
23:57:27 Join Xerion [0] (
23:57:37Bagderbut we can't be sure that will always happen
23:57:44Bagder209 seconds
23:57:52Bagderthat's pretty cool
23:57:53limbuswhoopsie: limbus.dyndns.org2900( 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0)
23:58:03Bagderthat's why then
23:58:05limbusthat explains the cpu :)
23:58:16amiconnSo as long as the whole build system doesn't break down to a single server in total, even a hanging one shouldn't cause problems
23:58:43Bagderamiconn: unless there are several builds that hang at once
23:58:44*petur hands 580 points to limbus

Previous day | Next day